Late Holocene Surface Ruptures on the Southern Wairarapa Fault, New Zealand: Link Between Earthquakes and the Uplifting of Beach Ridges on a Rocky Coast

Home , Wairarapa Fault

Late Holocene surface ruptures on the southern Wairarapa fault, New Zealand: Link between earthquakes and the uplifting of beach ridges on a rocky coast

T.A. Little1*, R. Van Dissen2†, E. Schermer3§, and R. Carne1# 1SCHOOL OF GEOGRAPHY, ENVIRONMENT AND EARTH SCIENCES, VICTORIA UNIVERSITY OF WELLINGTON, P.O. BOX 600, WELLINGTON 6140, NEW ZEALAND 2GNS SCIENCE, P.O. BOX 30368, LOWER HUTT 5010, NEW ZEALAND 3GEOLOGY DEPARTMENT, WESTERN WASHINGTON UNIVERSITY, MS9080, BELLINGHAM, WASHINGTON 98225, USA

ABSTRACT

The Holocene beach ridges at Turakirae Head, New Zealand, are remarkable because the fault that caused their uplift is accessible to paleoseismic trenching. Based on 40 14C samples from eight trenches, we identify ﬁ ve surface-rupturing earthquakes since ca. 5.2 ka (mean earthquake recurrence of 1230 ± 190 yr). The paleoearthquake record includes two more events than were recorded by the uplift and stranding of beach ridges at Turakirae Head. We conclude that beach ridges may provide an incomplete record of paleoearthquakes on oblique-reverse faults. The southern end of the Wairarapa fault includes several splays in the near surface at variable distances from Turakirae Head. Variable partitioning of slip between these splays (and perhaps the subduction interface down-dip of them) is inferred to have caused variable magnitudes of coseismic uplift at the coast, where at least one <3 m throw is not recorded by preservation of a ridge. Variations in wave climate or sediment supply (or interseismic subsidence) may also inﬂ uence the number of beach ridges preserved by governing the morphology of the storm berm and controlling its extent of landward retreat. Such retreat may cause a berm to overwhelm, or amalgamate with, the next-highest beach ridge, resulting in the omission of one ridge, as probably happened at Turakirae Head at least once. Our 14C data support the view that a widespread post–Last Glacial Maximum aggradational terrace in southern North Island, New Zealand, was abandoned soon after 12.1 cal yr B.P. From this, we infer that the Wairarapa fault has a late Quaternary slip rate of 11 ± 3 mm/yr.

LITHOSPHERE; v. 1; no. 1; p. 4–28; Data Repository item 2009053. doi: 10.1130/L7.1

INTRODUCTION in uplift of the hanging wall of that dextral- a series of similarly expressed earthquakes that reverse fault near the southern coast of the North have ruptured the nearby Wairarapa fault during Slip on reverse (or oblique-reverse) faults dur- Island (Fig. 1A) and in the generation of a set the past ~7000 yr (e.g., Wellman, 1969; Moore, ing large earthquakes may be accompanied by a of tsunami waves up to ~9 m high (Grapes and 1987; Hull and McSaveney, 1996). signal of coseismic uplift (or subsidence) near Downes, 1997). The coseismic uplift reached a Globally, the uplifted Holocene beach ridges the coast that may be preserved in the geological maximum near Turakirae Head, where the pre- at Turakirae Head are remarkable because the record (Atwater, 1987; Berryman, 1993; Wilson 1855 storm beach ridge was raised by as much fault that caused their uplift does not lie sub- et al., 2007a). Repeated surface-rupturing earth- as 6.4 m (Begg and Mazengarb, 1996; Hull and merged offshore but is exposed on land nearby quakes have the potential to generate a suite of McSaveney, 1996; McSaveney et al., 2006). and is accessible to paleoseismic study (e.g., uplifted coastal beach ridges that faithfully record This fossil beach ridge is today preserved as the Ota and Yamaguchi, 2004). In this paper, the sequence of earthquakes on that fault. Such youngest of at least four tectonically uplifted we determine the ages of surface-rupturing a paleoseismically advantageous situation might beaches on the headland (Fig. 1B) (Aston, earthquakes on the southern Wairarapa fault be most likely where these earthquakes have all 1912; Wellman, 1969; McSaveney et al., 2006). using fault-trenching techniques and 14C dat- been similarly large, involving ruptures of similar In addition to its large magnitude, the histori- ing. Undertaken on the same part of the fault dimensions, slip, and—especially—uplift; where cally well-documented 1855 earthquake was known to have ruptured in 1855, and under- coastal conditions have been continuously favor- remarkable for inﬂ uencing Charles Lyell (1868) pinned by more than 40 14C analyses collected able for the formation of beach ridges; and where to argue that earthquakes were associated with in eight trenches, our paleoseismic data allow the preservation potential of the abandoned land- vertical earth movements and slip on fault planes us to construct a comprehensive late Holocene forms has remained steadfastly high. (Grapes and Downes, 1997; Sibson, 2006), and earthquake chronology for the Wairarapa fault, New Zealand’s largest historic earthquake, the for its extremely large coseismic strike slip and to compare it with the published ages of Ms ~8.2 Wairarapa fault event in 1855, resulted (locally as high as ~18.5 m; Rodgers and Little, beach ridges at Turakirae Head. We are thus 2006). Since 1855, it has often been assumed that able to assess the completeness of that geo- the 1855 earthquake provided an analogue for morphic record and (potentially) the repeat- *Corresponding author e-mail: timothy.little@vuw. the style of deformation accompanying previous ability of large magnitude coseismic uplifts ac.nz. †[email protected]. earthquakes on this fault (e.g., Grapes, 1999). A on the fault. Although other studies have §[email protected]. corollary is that the uplifted gravel beach ridges compared beach ridge uplifts to the timing of #[email protected]. at Turakirae Head provide a complete record of historic earthquakes as much as ~500 yr ago

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 Late Holocene surface ruptures on the southern Wairarapa fault | RESEARCH

NIDFB

Hikurangi Trench Alfredton 50 Australian mm/yr Plate Mauriceville Northern 40° Tea Alfredton FaultSection 38 mm/yr Creek ui F. Alpine on ok Fault Hikurangi M Carterton Fault subduction zone Masterton F. 36 mm/yr Pacific Masterton Plate 41°S 170° 180° U Kaumingi Fault Waiohine River X D ult Wairarapa Fault Fa on Featherston gt n Pigeon Pacific Ocean li Wairarapa Valley el W Rimutaka Range Bush Figure 1. (A) Tectonic index map Ohariu Fault CentralCentral Coast Ranges U Rongotai Isthmus Cross showing major active faults and Fig. A other structures of the southern Wellington D Creek SectionSection Fault trenching North Island, New Zealand (largely Riverslea after Barnes, 2005; Begg and John- site (this study) ston, 2000; Lee and Begg, 2002), and Rimutaka location of sites along the Wairarapa Lake Kohangapiripiri Anticline SouthernSouthern SectionSection Fig. 2 Other localities fault where paleoseismic data have Orongaronga R. mentioned in text been collected previously and dur- Palliser Bay Turakirae ing this study (large open circles). Pliocene-Pleistocene strata Head Wharekauhau Smaller rectangles show location of Thrust 01020 detailed Wairarapa fault maps of Fig- ure A and Figure 2. Cross section X–X′ 176°E 175°E kilometers is presented in Figure 11B. Inset on Cape Palliser X’ A upper left shows plate-tectonic setting of New Zealand (plate motions taken from DeMets et al., 1990, 1994). NIDFB—North Island Dextral Fault Belt. (B) Oblique aerial photograph of uplifted beach ridges and Holocene wave-cut platform at Turakirae Head, view looking NW (photo by Lloyd Homer, GNS Science as annotated in McSaveney et al., 2006). The two uplifted Pleistocene marine terraces in the background were studied and assigned provisional ages by Ota et al. (1981). Distance between the two headlands is about 4 kilometers.

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 5

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

(e.g., Bookhagen et al., 2006; Ferranti et al., but this component is typically a small fraction inferred strike-slip fault is an apparent south- 2007), these studies typically suffer from the of the strike slip (<20%; e.g., Berryman, 1990; ward continuation of the main, central section of short time span of the historic data and from Heron et al., 1998; Rodgers and Little, 2006). In the Wairarapa fault (Begg and Johnston, 2000). ambiguities regarding the location of the sur- the offshore to the east and west of Wellington, Evidence for an along-strike continuation of this face rupture accompanying those historically upper-plate deformation is inferred to be domi- western strike-slip strand has not been found at felt earthquakes. To our knowledge, ours is the nantly contractional (e.g., Barnes et al., 1998, the coast to the west of Turakirae Head (Begg fi rst study attempting a one-to-one comparison 2002; Lamarche, 2005). and Johnston, 2000). Instead, that fault appears between uplifted strandlines and paleoseismi- Elastic dislocation modeling of global posi- to step southward onto a thrust segment (Muka cally documented fault ruptures over a time tioning system (GPS) data and seismicity data Muka fault) entering Palliser Bay just east of span of ~5 k.y. Our results have general impli- near Wellington suggest that the Hikurangi Turakirae Head (Begg and Johnston, 2000; Lit- cations for the sensitivity of uplifted gravel subduction zone occurs 20–25 km beneath the tle et al., 2008). The Rimutaka anticline to the beach ridges as a paleoseismic “tape-recorder” surface trace of these faults, and that this gently west is an active SW-plunging fold expressed by on exposed, high-energy coasts, and the vari- (~8°) west-dipping part of the plate interface the steep topography of the coastal ranges and ability of rupture styles on individual oblique- is currently locked and accumulating elas- by differential uplift of the Holocene wave-cut slip faults. Our results signifi cantly shorten tic strain (Reyners, 1998; Darby and Beavan, coastal platform near Turakirae Head (Ghani, estimates of the mean earthquake recurrence 2001; Wallace et al., 2004). It seems likely that 1978; Wellman, 1969; McSaveney et al., 2006). interval of the Wairarapa fault and increase North Island dextral faults merge downdip with Farther east, other structures at the southern end estimates for the late Quaternary slip rate of the underlying interface. The nature of interac- of the Wairarapa fault zone include Wharepapa this major fault. tion between the subducting plate interface and and the Battery Hill faults (Begg and Johnston, the intersecting North Island dextral faults is 2000), and an inferred blind thrust along the TECTONIC SETTING poorly understood. western side of Lake Onoke (Little et al., 2008). Seismic and bathymetric data suggest that at In central New Zealand, motion of the Pacifi c The Wairarapa Fault least two strands of the Wairarapa fault zone plate relative to the Australia plate occurs at continue offshore into Palliser Bay (Barnes and ~39 mm/yr in a direction of ~261° (DeMets et This NE-striking and steeply NW-dipping Audru, 1999; Barnes, 2005). al., 1990, 1994). The plate boundary in New Zea- dextral-reverse fault bounds the western fl ank The northern section of the Wairarapa fault land is characterized by subduction of oceanic of the Wairarapa basin, a trough of late Ceno- bifurcates eastward into a series of ENE-striking crust along the Hikurangi margin of the North zoic (mostly Pliocene-Pleistocene) marine and dextral-slip splays, such as the Carterton fault Island and oblique continental collision along terrestrial sedimentary strata in the forearc of (Begg and Johnston, 2000; Lee and Begg, 2002; the Southern Alps of the South Island (Fig. 1A). the Hikurangi subduction zone (Fig. 1A). The Langridge et al., 2005). The northern section is Faults in the southern part of the North Island southern end of the basin is 2.5–3 km thick well defi ned as far north as Mauriceville, beyond occur in a transition zone between these two where it is truncated against the fault (Hicks and which it links discontinuously into several more plate boundary zones. There, obliquely conver- Woodward, 1978; Cape et al., 1990). To the west, diffusely expressed, slower-slipping faults, such gent motion is partitioned between contraction- the adjacent Rimutaka Range is >900 m high, as the Alfredton fault (Schermer et al., 2004) dominated folding and thrust faulting in the sub- consisting of deeply eroded Mesozoic basement (Fig. 1A). Estimates of the late Quaternary slip merged accretionary wedge (e.g., Barnes and rocks. Based on the gross geometry of its trace, rate of the Wairarapa fault vary between ~7 and Mercier de Lepinay, 1997; Barnes et al., 1998) the Wairarapa fault can be divided into central, 12 mm/yr (Lensen and Vella, 1971; Van Dis- and in emergent parts of the forearc (e.g., Nicol southern, and northern sections (Fig. 1A). The sen and Berryman, 1996; Grapes, 1999; Wang et al., 2002, 2007), and strike-slip faulting (and central section consists of an en echelon array and Grapes, 2007), chiefl y because of uncer- vertical-axis fault-block rotations) in the onland of mostly left-stepping fault segments that are tainties in the age of the widely distributed and region farther to the west (e.g., Beanland, 1995; typically 1–2 km long (Fig. A in Appendix).1 recognized (Begg and Johnston, 2000; Lee and Beanland and Haines, 1998; Mouslopoulou et These discontinuous, dextral-oblique faults are Begg, 2002), but only sparsely dated, Last Gla- al., 2007; Van Dissen and Berryman, 1996; Wal- separated by contractional bulges or folds in cial Maximum aggradation surface referred to lace et al., 2004). In the onshore region near Wel- the area of their overlap that cause warping of locally as the “Waiohine” terrace. lington, the margin-parallel component of plate alluvial terrace surfaces (Grapes and Wellman, motion (up to ~18 mm/yr or ~70% of the total) 1988; Rodgers and Little, 2006). 1855 Earthquake and Raised Beach is mostly accommodated by dextral-slip on the The structurally complex southern section Ridges at Turakirae Head Wairarapa, Wellington, Ohariu, and Shepard’s of the fault (Fig. 2) includes the west-dipping Gully faults. These are the southernmost ele- Wharekauhau thrust and other active and inac- The Wairarapa fault northeast of Wellington, ments of the North Island Dextral Fault Belt, a tive fault splays. The Wharekauhau thrust is New Zealand, ruptured on 23 January 1855, belt of upper-plate strike-slip faults in the cen- locally overlapped by undeformed late Quater- resulting in ground shaking landslides, regional tral part of the North Island (Beanland, 1995). nary–Holocene gravels; however, some strands uplift, tsunamis, and a surface rupture that broke The faults probably initiated ca. 2 Ma or more of the Wharekauhau fault system show recent all three of the aforementioned sections of the recently by reactivation of preexisting reverse motion (Little et al., 2008). To the west, an fault. The rupture was up to 120 km long on land faults, and they have since accommodated at and at most 40 km long on the seafl oor to the SW most ~20 km of cumulative dextral slip (Bean- (Grapes and Downes, 1999; Rodgers and Little, 1 land, 1995; Begg and Mazengarb, 1996; Kelsey GSA Data Repository Item 2009053, Appendices 2006). At Ms ~8.2, it was the largest earthquake A–H and Figures A–E, is available at www.geosoci- and Cashman, 1995; Nicol et al., 2007). At the ety.org/pubs/ft2009.htm, or on request from editing@ in modern New Zealand history. South of Feath- surface, the faults dip steeply (typically to the geosociety.org, Documents Secretary, GSA, P.O. Box erston, on a 16-km-long part of the central sec- NW). Their dip slip is mostly up-to-the-west, 9140, Boulder, CO 80301-9140, USA. tion of the rupture trace (Fig. A), Rodgers and

6 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

26802680000000 26902690000000 Bu g same name). Farther west, based on the eleva- Hill tion difference between BR-1 and BR-2, the 1855 uplift reached a maximum of ~6.4 m at e the crest of the Rimutaka anticline, ~3 km NE g lt irarapa of Turakirae Head (McSaveney et al., 2006). n u a a R F The 1855 uplift decayed westward to ~1.5 m a ka p near Wellington (Grapes and Downes, 1999). a ra Lake Wa t a . In addition to BR-1 and BR-2, Turakirae Head 59905990000000 u ir F m a l i WairarapaW Fault il hosts at least three other higher and older, RimutakaR Range MataruaM H a t y raised beach ridges: BR-3, BR-4, and BR-5 a r ru a te (Wellman, 1969). McSaveney et al. (2006) t ManganuiM a radiocarbon dated BR-3 (2380–2060 cal. yr a BatteryB Hill F. n g B.P.) and BR-5 (6920–6610 cal. yr B.P.) and a

B Battery Battery u a i used the elevation difference between adjacent

e t t n e ridges as a measure of the coseismic uplift i r

l y c i that caused the stranding of the upper ridge. In t WharepapaW n h this way, they inferred an average incremental a a a r a e p Riverslea Site (coseismic) uplift of 7.3 ± 0.9 m at the crest of k a

a p

a the anticline and a mean Holocene uplift rate t u WharekauhauW Holocene, undiff. of 3.5 ± 0.02 mm/yr at that location. Based on m h i a re the premise that every Wairarapa fault earth- R k a D u h Last Glacial alluvial quake since ca. 7 ka has caused preservation of ? a u a corresponding beach ridge at Turakirae Head, ZoneZone ofof activeactive Last Interglacial marine they calculated a mean recurrence interval for 59805980000000 U D foldingfolding Wairarapa fault earthquakes of ~2200 yr. Early Quaternary For this study, eight paleoseismic trenches Lake were excavated across three different sites, and Wharekauhau Onoke Torlesse graywacke 40 samples were submitted for 14C dating. See thrust (inactive) D U Stream Appendix A for further information about our Muka Muka + younger Thrust fault surveying and trenching methods. Next, we will fault (approx. strike-slip Wharepapa F Strike-slip fault discuss our paleoseismic results at the Pigeon located) strands (inactive) Bush locality, then those at the Riverslea, and (active, near-vertical) Inactive TurakiraeMukamuka Head? or buried fault ﬁ nally those at Cross Creek (Fig. 1A). ~ 3 km ? 5 km F. (approx Anticline axis RESULTS Canyon located)apa Figure 2. Active fault traces and folds along the Wharekauhau thrust along the southern section of Pigeon Bush Site the Wairarapa fault zone (after Begg and Johnston, 2000; revisions by Little et al., 2008). Grid marks (in meters) refer to the New Zealand Map Grid Coordinate System. Site Geology and Trench Stratigraphy At the Pigeon Bush site (Figs. 1A and A), Grapes and Wellman (1988) interpreted two Little (2006) mapped displaced landforms, ing of the vertical component of the coseismic beheaded stream channels as evidence of the mostly small beheaded or offset stream chan- motion, although it is not a uniquely determined repeated dextral offset of a small stream gully nels, and inferred an average single-event strike aspect of the model (Darby and Beanland, 1992; crossing the Wairarapa fault (the last in 1855). slip (in 1855) of 15.5 ± 1.4 m. There were no Beaven and Darby, 2005). The two channels are abruptly and orthogonally contemporary observations of coseismic strike A ﬂ ight of uplifted gravel beach ridges truncated at the fault on its SE side. On the NW slip in 1855. Three of the 16 sites yielded strike- occurs at Turakirae Head on the exposed and side of the fault, a small, entrenched, and still- slip estimates of 13–14.5 m for the penultimate rocky southern coast of North Island, New Zea- active source gully to these channels is similarly earthquake, and several yielded estimates of land. There, the crest of the modern storm berm linear and fault-transverse (Fig. 3). This geo- the throw in 1855 (up-to-the-NW). At ~2–3 m, (BR-1) is variably located 2–7 m above mean morphology implies that the two paleochan- these throws agree with contemporary obser- sea level, where it presumably marks the runup nels were beheaded as a result of two consecu- vations of the height of the scarp in the 1855 limit of present-day storm waves. The next tive earthquakes, and there is no evidence for a earthquake (Grapes and Downes, 1999). Based highest beach ridge (BR-2) was abandoned as a temporary phase of stream diversion parallel to on the unusually large ratio of displacement to result of uplift in 1855 (Begg and McSaveney, the scarp (i.e., for a third or fourth slip incre- rupture length (D/L) for this earthquake, Rod- 2005; McSaveney et al., 2006). Contemporary ment; Rodgers and Little, 2006). These authors gers and Little (2006) argued that the rupture observations (mostly by a surveyor; Roberts, measured 18.7 ± 1.0 m of dextral slip and ≥1.25 extended several tens of kilometers downdip 1855) and other geological data indicate that ± 0.5 m of vertical slip for the younger, aban- (W) to merge with, and co-rupture, a down-dip the 1855 uplift was approximately zero on the doned channel relative to the upstream gully and part of the subduction interface. This conclusion shore of Palliser Bay, to the east of the Muka 32.7 ± 1.0 m and ≥2.25 ± 0.5 m of slip (respec- is consistent with elastic dislocation model- Muka rocks (near the trace of the fault of the tively) for the older abandoned channel. This

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 7

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

27013602701380 2701400 2701480 Pigeon Bush

2701440

6004940 tilted Waiohine Figure 3. Microtopographic map of the Pigeon Bush stream offset surface site showing location of beheaded 6004920 ac stream channels and the Pigeon U tiv fence line e Bush 1 and 2 trenches (PB-1, PB-2, st re lt am respectively). Base map is from u D a pit ? Rodgers and Little (2006) and was F surveyed using a combination of a 6004900 p real-time kinematic differential ra global positioning system (GPS) a pit ir methods employing a differential a PB-1 ? correction that was tied to a base WairarapaW Fault station at benchmark LUCENA (NZMS270-S27A) and 3-D laser- 6004880 PB-2 ranging techniques. Trench corners were located by real-time kine- ? matic GPS survey tied to the same base station. Grid marks (in meters) North Wang & Grapes (2007) refer to the New Zealand Map Grid O.S.L. sample Coordinate System. M.S.L.—mean 6004860 ? (”PB-1,” 7.0 ± 0.5 ka) sea level; O.S.L.—optically stimulated luminescence. 0 20m

Wang & Grapes (2007) 6004840 O.S.L. sample (”PB-2,” 4.3 ± 0.5 ka) Contour Interval = 25 cm (elevations relative to M.S.L.)

larger offset suggests that the older beheaded section along the SW wall of the same trench collapse of the channel banks (e.g., units Cg1 channel had previously been displaced by 14.0 (Fig. 4B). The latter wall exposed a richer to Cg7 in Fig. 4B). ± 1.0 m of dextral slip and ~1.0 m of vertical slip and more easily differentiable stratigraphic The fl uvial part of the channel-infi ll sequence prior to incision of the younger one. sequence and was logged in detail (Fig. 4B). In contains abundant detrital charcoal fragments up The fault displaces the “Waiohine” terrace both of the Pigeon Bush trenches, fl uvial terrace to 3 cm in diameter, and these are locally inter- gravels and forms a steep, SE-facing scarp gravels at the base of the trench (e.g., units Tgr, bedded with an organic-rich paleosol (unit Ps). that is ~6 m high. On the northern side of this Tgr-1, and Tgr-2) are overlain by up to ~75 cm Four charcoal samples from the channel infi ll scarp, the uplifted Waiohine terrace is tilted SW, of silt (see unit Si in Figs. 4A and 4C). This ter- were radiocarbon dated. The fi rst three (PB-1, whereas on its southern side, it is partly buried race covering layer of silt is probably what was PB-2, PB-3) yielded ages of 649–497 cal yr B.P. beneath younger deposits. On the SE side of OSL dated by Wang and Grapes (2007). The (Table 1). These overlap (at 95% confi dence) the fault, an ~1-m-thick layer of silt (unit Si) aforementioned terrace units were incised to a with each other and with two charcoal samples mantles the terrace gravel. This layer was later depth up to ~2 m by the younger stream channel (DR0425G and DR0425J) collected and dated incised by the two channels (Figs. 4A and 4C). (Fig. 4A). None of these stream-incised terrace from a nearby pit by Rodgers and Little (2006). Wang and Grapes (2007) dated two samples of units yielded dateable organic material. The All six samples are apparently derived from a the silt cover bed by optically stimulated lumi- scoured unconformity defi ning the base of the single population of charcoals delivered into nescence (OSL) methods (Fig. 3) and obtained channel incision is mostly overlain by bedded the channel by a stream that aggraded rap- ages of 7.0 ± 0.5 ka and 4.3 ± 0.5 ka. gravel and sand, including units Sgr and Fgr-1 idly(?) after a bush fi re. Based on the age of the Trench PB-1 was cut orthogonally across to Fgr-5 in Figure 4B. These units are compact youngest of these samples (PB-1), the fi re prob- the younger of the two abandoned channels and well sorted, with rounded to subrounded ably took place at, or soon after, ca. 546–497 (Figs. 4A and 4B; Appendix B). Due to a mean- clasts, and they are interpreted to be axially cal yr B.P. (1404–1453 cal. yr A.D.), perhaps der in the channel (see Fig. 3), it was excavated transported fl uvial deposits. The remaining in response to Maori burning. Another sample, in a nearly orthogonal (and therefore narrow) channel-infi lling units are interpreted, on the PB-5, was collected in the modern soil, 32 cm section along the NE wall of PB-1 (Fig. 4A) basis of their poor sorting and clast subangu- below the ground, and it yielded an age of 490– and in a more oblique (and therefore wider) larity, to be colluvial units derived by lateral 315 cal yr B.P.

8 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

Poorly sorted channel-infilling gravel and silt (undiff.). NW SE CfuCfu Interpreted as a mixture of axially delivered fluvial deposits and bank-derived colluvial deposits. A NE wall, Pigeon Bush 1 Trench (PB-1)– youngest abandoned channel Si Yellowish-gray silt (locally fine pebbly) 3 SSii

ground surf SSii ace

t p-soil opogra xis se of tfo topsoil SgrSgr Lens of very poorly sorted silty pebble gravel & pebbly silt phic channel a ba se o SiSi Si 2 SgrSgr CCfufu TgrTgr Crudely bedded fluvial terrace gravels 1 scoured TgrTgr edge of ch excavation Vertical Distance (meters) Distance Vertical en channel base of tr 0 9 8765 4 321 0 Horizontal Distance (meters) GrsGrs Soil-modified gravelly silt

Orange-mottled paleosoil, with abundant charcoal Ps (developed on pebbly channel deposits) SE NW Pebble-granule gravels (subrounded-rounded clasts). SW wall, Pigeon Bush 1 Trench (PB-1)– youngest abandoned channel FgrFgr Interpreted as fluvial channel deposit. Five bodies n are labelled individually (n = 1-5). B xis PB-5 nnel a u topsoil t ic cha Cgs-3 Matrix-supported gravelly silt and silty gravel. Interpreted gro nd surface opograph 3 as chiefly colluvium, but may include some waterlain deposits. GGrsrs CgnCgn Cg7Cg7 4 Cg5 Seven bodies are labelled individually (n = 1-7). Cg6Cg6 FFgrgr Cg1Cg1 4 Ps Cg2Cg2 1 Ps FgrFgr 5 1 2 4 Cg4 Silty pebble gravel (subrounded clasts). SSgrgr FgrFgr 3 FgrFgr sisi SgrSgr SSi SSi 2 Interpreted as channel-infilling fluvial deposit Tgr-2Tgr-2 scoured Tgr-2 Tgr-1Tgr-1 PB-3 ion at edge of av sisi PB-2 xc Si Massive silt (locally fine pebbly) PB-1 ch e channel base of tren 1

Tgr-2TTgr-2gr-2 Terrace gravels (pebbly with silty matrix) Vertical Distance (meters) Distance Vertical

0 1234 5 678 9Tgr-1Tgrr--1 Crudely bedded terrace gravels (cobbly with sandy matrix) Horizontal Distance (meters)

NESE wall, Pigeon Bush 2 Trench (PB-2)– older abandoned channel SW C tpstps Topsoil PB-22PB-22 nd surface Si grou 2 to SiSi pographic channel axis Gravelly silt (>60% matrix). Clasts are angular-subangular. CdfCdf tpstps SiSi Unit is interpreted as channel-infilling debris flow.

1 CdfCdf Well-sorted yellowish-gray silt scoured TgrTgr Si containing sparse fine pebbles. tion edge of h excava base of trenc 0 channel Large cobble Vertical Distance (meters) Distance Vertical TgrTgr Crudely bedded terrace gravels (fluvial) -1 0123 4 567 8 Horizontal Distance (meters) PB-2PB-22 Radiocarbon sample (charcoal). See Table 1. Figure 4. Logs of the Pigeon Bush trenches. (A) NE wall of Pigeon Bush 1 (PB-1) trench across the younger of the two abandoned stream channels; (B) SW wall of the same trench; (C) SE wall of the Pigeon Bush 2 (PB-2) trench across the older of the two abandoned stream channels. See Figure 3 for location of the trenches. See Table 1 for 14C analytical results of the numbered samples. The unit labels and brief unit descriptions for each trench pertain only to that particular trench wall and are not meant to apply necessarily to any other trench log. See Appendix B for detailed stratigraphic descriptions of the logged units.

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 9

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

## 20 (5.6) (5.6) 20 Mode 1 Mode 2 Mode 3 Calibrated calendar yr B.P. at 95.4% confidence 95.4% at B.P. yr calendar Calibrated (% in mode) (% in mode) (% in mode) (% in C ages using liquid scintillation counting, except Wk21132, Wk21143, except Wk21132, Wk21143, counting, scintillation liquid C ages using

14 %Modern 68.7 ± 0.368.7 (2.1) 3320–3290 (93.3) 3270–2970 57.4 ± 0.357.4 (85.0) 5070–4850 (10.4) 5280–5170 sphere atmospheric data from McCormack et al. (2004). (2004). McCormack et al. from data sphere atmospheric

2 ± 0.3 58.3 (77.4) 4980–4800 (14.4) 4770–4690 (3.6) 4680–4640 2 ± 0.3 58.4 (69.4) 4970–4800 (20.3) 4770–4690 4680–46 2 ± 0.3 56.4 (2.4) 5450–5410 (93.0) 5330–5040 §§ 0.2 74.3 ± 0.4 (13.7) (81.7) ± 2250–2150 2490–2300 74.3 0.2 0.2 0.2 ± 0.3 58.4 (67.8) 4970–4790 (27.6) 4770–4620

C 13 fractionation. Quoted errors are one standard deviation due to counting statistics statistics counting to due deviation one standard are Quoted errors fractionation. 5 ± 0.2 ± 0.2 5 ± 0.5 75.1 (95.4) 2350–2130 ± 0.2 3 ± 0.3 65.5 (95.4) 3690–3460 δ 9.3 ± 0.2 ± 0.2 9.3 ± 0.2 24.9 (95.4) 13,160–12,900 (‰, PDB) –28.5 ± 0.2 ± 0.2 –28.5 ± 0.4 71.3 ± 0.2 –27.4 (95.4) 2870–2730 ± 0.3 72.0 (58.6) 2780–2680 (12.6) 2650–2610 (24.2) 2600–2409 –28.0 ± 0.2 ± 0.2 –28.0 ± 0.3 57.1 (95.4) 5290–4870

††

41 ± 0.2 –27.6 ± 0.3 63.2 (95.4) 4090–3830 42 ± 0.2 –28.2 ± 0.3 61.4 (95.4) 4420–4140 ) σ ± 37 ± ± 0.2 –28.8 ± 0.4 87.8 (95.4) 970–800 ± 36 ± ± 0.2 –28.3 ± 0.4 88.8 (95.4) 920–740 (±1 (yr B.P.) Analytical age (cm) Depth Northing NZ Map Grid , AMS); Wk—Waikato Radiocarbon Dating Laboratory (conventional (conventional Laboratory Radiocarbon Dating Wk—Waikato , AMS); 2689117 6001632 140 42 3980 ± ± 0.2 –28.6 ± 0.3 60.9 (7.7) 4520–4460 (86.6) 4450–4230 (1.1) 4200–4180 Easting** Easting** TABLE 1. WAIRARAPA FAULT TRENCHING RADIOCARBON DATING RESULTS DATING RESULTS RADIOCARBON FAULT TRENCHING WAIRARAPA TABLE 1. charcoal Material NZ Map Grid Map NZ Material # Bush; RVL—Riverslea; CC—Cross Creek. Bush; RVL—Riverslea; Log unit Log § Trench † Conventional age as per Stuiver and Polach (1977). This is based on the Libby half-life of 5568 yr with correction for isotopic for correction yr with 5568 of Libby half-life the on This is based (1977). Polach and Stuiver as per age Conventional Hemi Southern (2005) incorporating of Ramsey v. 3.10 OxCal using 1950 years before to calendar were calibrated Radiocarbon ages Parts per thousand with respect to Peedee belemnite (PDB). respect to Peedee belemnite (PDB). with Parts thousand per Abbreviations for trench names: PB—Pigeon trench for Abbreviations NZA—Rafter Radiocarbon Laboratory (Accelerator Mass Spectrometry Mass (Accelerator Laboratory Radiocarbon NZA—Rafter units. of the logged descriptions detailed for F and D, B, C, Appendices See † § †† §§ ## #

**Grid references are ±2 m precision. precision. ±2 m are references **Grid multiplied by an experimentally determined laboratory error multiplier of 1. 1. of multiplier error laboratory determined experimentally by an multiplied CC-2-31 NZA26178 CC-2 osi-2 charcoal 2698114 6001633 112 2167 ±30 ± 0.2 –26.6 ± 0.3 75.8 (4.1) 2294–2267 (90.9) 2154–1991

PB-1 PB-3 NZA26101 PB-1 NZA26103 PB-1 Cg2 3 charcoal CC-1-7 2701390 charcoal Wk19148 6004889 2701389 CC-1 98 6004892 30 528 ± pt5 83 ± 0.2 –25.8 ± 0.3 93.0 org. clay 30 557 ± (96.5) 546–497 ± 0.2 –25.2 2698166 ± 0.3 92.7 6001616 (95.7) 554–504 90 39 2385 ± ± –28.7 CC-3-6 Wk21133 CC-3 pt3 org. silt & silt CC-3 pt3 CC-3-6 Wk21133 org.

CC-1-9 Wk19150 CC-1 pt4 org. clay 2698166 6001616 126 2723 ± 40 RVL-2 RVL-3 RVL-4 –28. ± NZA26180 51 CC-1-1a-I NZA26105 RVL-2 CC-1-1a-II Wk191161 Wk191160 NZA26122 CC-1 2723 RVL-2 CC-1-6 ± 126 CC-1 6001616 RVL-2 2698166 clay si-15a CC-1 pt6 cgvl-16 pt5 CC-1-8 Wk19149 Wk191162 CC-1 clay 2302 pt4 2698167 CC-1-9 Wk19150 peat charcoal si-9b 6001616 110 org. clay pt6 6001616 CC-1 47 charcoal 2698166 2698166 CC-1-10 956 wood 6001616 org. 26 1047 undiff. CC-1 CC-1-11 Wk19152 wood fibers 38 2688330 org. undiff. CC-1 –29. pt2 Wk19151 CC-1-12 Wk19153 2688330 2698168 41 tree pt, 2688330 CC-1 pt1 CC-1-13 Wk19154 ± peat 63 6001615 peat –2 CC-1 pt, 5985821 CC-1 CC-1-14 Wk19155 ± 5985821 ± peat 233 2698166 CC-2-30 5985821 6001616 2635 190 11,173 wood 2698166 49 6001616 3394 186 2698168 41 pt4 Wk19156 270 30 530 ± 6001615 2698167 CC-2 30 117 ± 30 906 ± ± 0.2 –28.0 peat ± 0.2 233 –27.0 ± 0.2 –27.3 6001616 ± 0.3 93.0 ± 0.4 97.9 ± 0.3 88.7 pt ± 57 10,503 (97.0) 547–497 118 ± 0.2 –28.6 2698166 (17.2) 264–220 (6.6) 897–873 ± 0.2 27.2 37 (95.4) 2254 ± 12,700–12,100 org. clay 6001616 ± 0.2 –28.3 146–0 (81.2) (84.7) 803–722 ± 0.3 75.5 2698113 176 (95.4) 2340–2110 40 3018 ± 6001634 ± 0.2 –28.5 153 43 4331 ± ± 0. –27.2 CC-2-37 CC-2-38 NZA26019 CC-3-2 NZA26121 CC-2 CC-2 Wk21132 co-2 CC-3 osi-1 wood pt4 wood charcoal 2698113 2698114 2689114 6001634 6001633 142 6001635 121 4759 ±30 128 40 4424 ± ± 0.2 –24.8 29 2132 ± ± 0.2 –25.1 ± 0.2 54.9 ± 0.3 57.3 ± 0.2 –28.4 (22.6) 5580–5504 ± 0.3 76.7 (93.5) 5046–4842 (75.9) 5486–5316 (95.4) 2150–1940 (0.9) 5209–5200 CC-4-10 Wk21142 CC-4 pt, undiff. peat 2698162 6001607 83 43 4320 ± ± –27.8 CC-4-13 Wk21144 Auger 3 CC-4 Wk21137 CC-Auger basal peat si wood 2698140 wood 6001610 2698162 273 6001608 43 4714 ± 56 ± 0.2 –26.3 ± 0.3 55.6 30 3058 ± (8.2) 5580–5530 ± 0.2 –25.8 (87.2) 5490–5300 ± 0.3 68.3 (95.4) 3340–3070 DR0425 G NZA20657 pit N/A charcoal 2701378 6004903 154 35 573 ± ± 0.2 –25.1 ± 0.4 92.5 (98.2) 649–523

Wk21144, which are AMS ages by Rafter Laboratory). Laboratory). Rafter ages by AMS are Wk21144, which CC-2-33 CC-2-34 NZA26018 Wk19163 CC-2 CC-2 osi-1 pt wood org. clay 2698114 2698113 6001632 6001634 118 167 4583 ±40 41 4320 ± ± 0.2 –25.5 ± 0.3 56.1 ± 0. –26.5 (95.1) 5316–5035 (1.7) 5005–4980 CC-4-2 Wk21138 CC-4 pt, undiff. peat 2698164 6001609 c. 175 3688 ± ± ± 3688 43 3914 175 160 ± c. undiff. peat 6001609 CC-3-E 2698164 c. undiff. peat 6001609 4500 CC-3-F 2698164 undiff. peat 130 pt, 6001608 CC-3-L 2698161 Wk21134 CC-4 pt, CC-4-2 Wk21138 CC-4 Wk21135 pt, CC-4-3 Wk21139 CC-3 CC-4 Wk21136 CC-4-4 Wk21140 CC-3 CC-3 pt pt pt org clay org clay 2689113 peat 2689113 6001633 6001633 2689113 190 168 6001633 42 3856 ± 40 2892 ± ± 0.2 –28.6 238 ± 0.3 61.9 ± 0.2 –28.4 39 4455 ± ± 0.3 69.8 (91.4) 4410–4070 (94.0) 3080–2840 ± 0.2 –29.3 (4.0) 4040–3990 (1.4) 2820–2790 DR0425 J NZA20658 PB-2 pit PB-5 PB-22 NZA26102 RV-1 N/A PB-1 NZA26179 NZA26104 PB-1 NZA26119 PB-2 charcoal Cg2 RVL-1 Grs 2701378 charcoal Cdf ss-3 6004903 charcoal 2701390 charcoal charc 15 2701389 6004889 2701378 30 603 ± 6004892 73 2688314 2004886 ± 0.2 –26.0 ± 0.4 92.2 32 5985806 30 595 ± 44 ± 0.2 –27.2 (98.7) 655–538 164 25 380 ± ± 0.3 92.2 30 521 ± 30 1414 ± ± 0.2 –24.1 ± 0.2 –27.6 (18.7) 627–602 ± 0.3 94.7 ± 0.2 –24.9 ± 0.3 93.1 ± 0.3 83.3 (99.2) 490–315 (73.9) 563–518 (95.9) 543–495 (80.0) 1312–1260 CC-2-35 Wk19158 CC-2 pt wood 2698114 6001633 137 43 4608 ± ± 0.2 –24.7 ± 0.3 56.3 (4.1) 5450–5410 (91.3) 5330–5040 CC-4-6 Wk21141 CC-4 pt, undiff. peat 2698161 6001608 111 43 4600 ± ± 0. –27.8 CC-4-11 Wk21143 CC-4-16 CC-4 Wk21146 *The first two samples in this table are from (2006). and Rodgers Little CC-4 cwA cwB wood org. clay 2698161 2698161 6001608 6001608 110 76 43 4764 ± 42 4289 ± ± 0.2 –26.7 ± 0.3 55.3 ± 0.2 –28.3 ± 0.3 58.6 (27.5) 5590–5500 (41.6) 4870–4780 (67.9) 5490–5310 (53.8) 4770–4610 Field no.* Field no.* Lab no.

10 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

TABLE 2. AGE RANGES FOR PALEOEARTHQUAKES ON THE SOUTHERN WAIRARAPA FAULT, BROKEN DOWN INDIVIDUALLY BY TRENCHES AND SITES Site (trenches) Local 14C-derived maximum age range Preferred age Historical constraints event (cal. yr B.P., 95% confidence)* (cal. yr B.P., 95% (if any) Younger limit (sample no.) Older limit (sample no.) confidence) (cal. yr B.P., 95% confidence) Pigeon Bush (PB-1, PB-2)

Trench PB-1 Pb1 N/A 490–315 (PB-5) 95 (=1855 A.D.) 95 (=1855 A.D.)

Trench PB-1 Pb2 546–497 (PB-1) N/A Riverslea (RV-1, RV-2)

Trench RVL-2 Rv1 264–0 (RVL-3) 547–497 (RVL-2) 95 (=1855 A.D.) 95 (=1855 A.D.) † Trench RVL-2 Rv2 547–497 (RVL-2) 897–722 (RVL-4) Cross Creek South (CC-1, CC-4)

Trench CC-1 CCS1 N/A 970–800 (CC-1-14) 95 (=1855 A.D.) 95 (=1855 A.D.)

Trench CC-1 CCS2 970–800 (CC-1-14) 920–740 (CC-1-13)

Trench CC-1 CCS3 920–740 (CC-1-13) 2340–2110 (CC-1-6) 2340–2110 (CC-1-6)

Trench CC-4 CCS3 N/A 3340–3070 (CC-4-13)

Composite of CC-1 and CC-4 CCS3 920–740 (CC-1-13) 2340–2110 (CC-1-6)

Trench CC-1 CCS4 3320–2970 (CC-1-10) 3690–3460 (CC-1-12)

Trench CC-4 (NE wall) CCS4 N/A 3830–4090 (CC-4-2)

Trench CC-4 (SW wall) CCS4 3340–3070 (CC-4-13) 4870–4610 (CC-4-16)

Composite of CC-1 and CC-4 CCS4 3340–3070 (CC-4-13) 3690–3460 (CC-1-12)

Trench CC-4 CCS5 4970–4620 (CC-4-10) 5450–5040 (CC-4-6) Cross Creek North (CC-2, CC-3)

Trench CC-2 CCN1 N/A 2294–1991 (CC-2-31) 95 (=1855 A.D.)

Trench CC-3 CCN1 N/A 2150–1940 (CC-3-2) 95 (=1855 A.D.)

Composite of CC-2 and CC-3 CCN1 2150–1940 (CC-3-2) 95 (=1855 A.D.)

Trench CC-2 CCN2 N/A 2294–1991 (CC-2-31)

Trench CC-3 CCN2 N/A 2150–1940 (CC-3-2)

Composite of CC-2 and CC-3 CCN2 N/A 2150–1940 (CC-3-2)

Trench CC-2 CCN3 2294–1991 (CC-2-31) 4970–4620 (CC-2-34, also -30)

Trench CC-3 CCN3 2150–1940 (CC-3-2) 3080–2790 (CC-3-F)

Composite of CC-2 and CC-3 CCN3 2294–1991 (CC-2-31) 3080–2790 (CC-3-F)

CC-2 CCN4 4980–4640 (CC-2-30, also N/A 34)

CC-3 CCN4 4410–3990 (CC-3-E) 5280–4850 (CC-3-L)

Composite of CC-2 and CC-3 CCN4 4980–4640 (CC-2-30) 5280–4850 (CC-3-L) 5209–4842 (CC-2-38) Note: See text for further information about how individual earthquake events were identified and their ages were bracketed at each trench site. *Limiting 14C ages. Field number of event-constraining 14C sample is shown in brackets (see Table 1). Age highlighted in bold indicates limiting age (at 95% confidence) for the sample at that site that most narrowly constrains the event timing (either its maximum or minimum age limit). Where possible, this is based on a composite analysis of all trench data at that site. †Based on our preferred interpretation that the dated wood of sample RVL-4 was a detrital particle; if it was a root fragment, then its age would not necessarily define an older age limit.

Trench PB-2 was cut orthogonally across the detrital particles within the fl uvial deposits these scarps offset small gullies by 6–8 m dex- older of the two beheaded channels (Fig. 4C). that infi ll the incisional scour that defi nes the trally and 1–3 m in an up-to-the-NW sense Fluvial terrace gravels at the base are scoured youngest beheaded channel. This channel was (Little et al., 2008). The linearity of these traces beneath an ~1-m-deep, channel-bounding uncon- cut immediately after the penultimate earth- suggests a steeply dipping fault. The Riverslea formity. The channel is infi lled by a massive, quake, resulting in the unconformity between trenches are located on a late Holocene river ter- matrix-supported pebbly clay, which is inter- the “Tgr” and “Si” units and the younger units race just west of Manganui Stream, where Begg preted as a debris fl ow. No fl uvial deposits are above them that infi ll the scour (these are shown and Mazengarb (1996, p. 84) suggested that a present. A single piece of charcoal near the top of in gray patterns on Figs 4A and 4B). This chan- fault may have ruptured in 1855, as is consis- the debris fl ow (PB-22) yielded a 14C age of 543– nel was later displaced and abandoned as a tent with accounts of that earthquake’s ruptur- 495 cal. yr B.P., which suggests that it is another result of slip during the most recent earthquake ing southward to the coast (Grapes and Downes, element of the aforementioned burn population on the fault. Thus, our preferred age of the burn 1997). Trench RV-1 was excavated perpendicu- that had become entrained into the debris fl ow. event (ca. 546–497 cal yr B.P.) provides a mini- lar to the main ENE-striking, ~3-m-high topo- mum age constraint for the penultimate earth- graphic scarp on the terrace (Fig. 5). Trench

Interpretation of Surface-Rupturing Events quake at this site (event Pb2, Table 2). Historical RV-2 was placed slightly higher up this slope Each of the last two earthquakes on the Wair- data indicate that the youngest earthquake here to intersect a less conspicuous scarplet on the

arapa fault at this site resulted in abandonment (event Pb1) took place in 1855. uplifted, NW side of the main scarp. of a stream channel immediately downstream The strata in trench RV-1 exhibit no fault- of the narrow headwater gully and incision of a Riverslea Station Site ing to ~2 m depth, but they are apparently new channel in downstream continuity with that folded (Fig. 6A; Appendix C). Loose, unstable gully. Hoping to date the last two earthquakes, Site Geology and Trench Stratigraphy ﬂ uvial sands (units ss-1 to ss-6) and grav- we excavated stratigraphic trenches at right On both sides of Manganui Stream, series of els (units gvl-1 to gvl-4) prevented deeper angles to each of the two channels to date their NE-striking, discontinuous scarps traverse steep trenching. Most strata (excluding cross-beds) incision and abandonment. The only organic hillslopes underlain by undated early Quater- dip 20°–25° to the SE. The decimeter-scale material that we found (charcoal) occurs as nary(?) gravels (Figs. 1A and 2). In two places, cross-beds mostly indicate SE ﬂ ow subparallel

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 11

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL. 2688320

47 M RV-2 a 45 n 43 g a n

5985820 u

r e

fold scarp a ? m 2688260

? 41 D RV-1 ? ? 52 2688380 N small 45 tributary 5985760 to Manganui 20 m Stream Contour interval 0.5 m (M.S.L. datum)

Figure 5. Microtopographical map of the Riverslea fault trenching locality. Map is based on 1020 points that were surveyed using real-time kinematic differential global positioning system (GPS) methods. GPS survey was tied to an arbitrary local base station and was not corrected against a gazetted geodetic benchmark. Grid marks (in meters) refer to the New Zealand Map Grid Coordinate System. M.S.L.—mean sea level.

to the modern Manganui Stream. The basal scarp (Fig. 5). There, fl uvial sands, silts, and soil- and colluvium-infi lled fi ssure just beneath contacts of gravel bodies are channelized and channelized gravels similar to those in trench the topographic surface (Figs. 6C and 6D). exhibit a cut-and-fi ll relationship to surround- RV-1 are vertically offset by several steeply Charcoal from unit si-15a in the faulted walls of ing beds. We interpret the sequence to be tec- dipping faults (Fig. 6C; Appendix C). Sand the fi ssure yielded a 14C age of 547–497 cal. yr tonically tilted because: (1) the SE dip of the bed ss-7 is separated vertically in an up-to-the- B.P. (sample RVL-2, Table 1). Charcoal from in beds accords with the SE facing direction of NW sense by ~20 cm across fault strand 2, and the fi ssure infi ll (unit cgvl-16, sample RVL-3, the scarp; and (2) these dips appear too steep by ~50+ cm across fault strand 1 (assuming Table 1) yielded a 14C age of 0–264 cal yr B.P. to be primary. By analogy with the small and that either ss-2 or ss-3 is equivalent to ss-7). shallow morphology of channels in present- Near the base of the trench, fault strands 2 and Interpretation of Surface-Rupturing Events day Manganui Stream, it seems unlikely that 3 cut a thick silt unit (si-9a and si-9b). Both The two Riverslea trenches contain evidence

ﬂ uvial sediments, in particular parallel-lami- of these faults terminate upward against a cut- for two earthquakes (events Rv1 and Rv2 in nated sands (Fig. 6B), would be deposited at a and-ﬁ ll sequence of depositionally overlap- Table 2). In trench RV-2, the most recent earth-

primary dip of 23°–25° in beds that are several ping, younger gravel (unit gvl-18). Unit si-9b quake (Rv1) ruptured to the surface and cut meters long. Unfortunately, the trench was not contained a centimeter-diameter fragment of sediments younger than ca. 500 cal. yr B.P. to long enough to convincingly expose the base of fi brous wood that yielded a 14C age of 897–722 create the fi ssure. The fi ssure was infi lled by the main scarp to the SE where these beds pre- cal yr B.P. (sample RVL-4, Table 1). soil debris containing “modern” charcoal and sumably return to horizontal, although a small Abrupt truncations and thickness changes is thus consistent with this earthquake being lens of gravel and sand at the extreme SE end across these and other faults in the trench sug- the 1855 rupture. Based on the stratigraphic of the trench (unit gss) appears horizontal. A gest strike-slip motion. Units gvl-5b, and gvl- mismatches across fault strand 1, slip during sample of detrital charcoal from the sands and 17 and si-5 (which includes a distinctive sandy the fi ssure event is inferred to have been pri- fi ne gravel of unit ss-3 (sample RV-1, Table 1) interval) are truncated against the NW side of marily strike slip. In the same trench, evidence 14 yielded a C age of ca. 1300 cal yr B.P. fault strand 1 without any apparent correlatives for an older (penultimate) earthquake (Rv2) Trench RV-2 was excavated across the being found on the SE side of the fault. Fault includes the depositional overlap of fault 3 decimeter-high scarplet to the NW of the main strand 1 diverges upward into a wedge-shaped, (and some splays of fault 2) by undeformed

12 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

SE ss-4

A Riverslea trench 1 (RVL-1), SW wall gvl-1

fold scarp si gvl-2 gvl-4 si silty topsoil ss-1 csscss

ss-2 gss gvl-4 ss-5ss-5 ss-3 gvl-3 tion excava Fig.Fig. 66bb base of ss-6 RV-1 1312-1260 cal. yrs B.P. 1 m si silt & topsoil

Horizontal = ss-1 to 6 well-bedded and/or laminated fluvial sand 0.5 Vertical Distance (Meters) Distance Vertical Vertical Scale css, gss (ss-5 = finely parallel-laminated sand) gvl 1 to 4 terrace gravel (fluvial) 14 1 m 0.5 RV-1 C Sample, cal. yrs. B.P.

Photo: B Horizontal Distance (Meters) inclined laminations horizontal string line in fine sand D Photo: infilled C Riverslea 2 trench, fault fissure (RVL-2), NE Wall 5050 cmcm

RVL-2 NW RVL-3 547-497 cal. yrs B.P. 0-264 cal. yrs B. P. infilled fault fissure SE gvl-17 si silty topsoil

si-15a

cgvl-16 gvl-5b5b 5050 cmcm intervalinterval ww// si-15b sandsand sstringerstringers si-5 gvl-18 si-14c si silt scoured Fig. 6d channel base si-5 gvl-4 ss 4 ss-3 fluvial sand gvl-1 ss-14b 12 si-3 Vertical Distance (Meters) Distance Vertical

ss-12 gvl-11 gvl fluvial gravel si-14a gvl11 ss-2 12 gvl-10b cgvl colluvial gravel si-10 gvl-11 gvl-8 RVL-4 2 si-9b RVL-4 897-722 B.P. 14 si-9a C sample ion avat cal yr. B.P. exc 7 of ase 1 3 b Horizontal Distance (Meters) flt: 075/90 flt: 208/70 1 meter

gvl-6 ss-7 Figure 6. (A) Log of SW wall of Riverslea 1 trench (RVL-1) across folded scarp of Wharekauhau thrust near Manganui Stream. (B) Photograph of inclined (probably tectonically tilted) laminae in fi ne- to coarse-grained sand layer (unit ss-5 in RVL-1 trench). (C) Log of part of the NE wall of the nearby Riv- erslea 2 trench (RVL-2). (D) Photograph of infi lled fault fi ssure in RVL-2 trench. See Figures 2 and 5 for trench locations. See Table 1 for 14C analytical results of the numbered samples. See Appendix C for detailed stratigraphic description of the logged units. M.S.L.—mean sea level.

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 13

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

channel gravel (gvl-18). A sample of ﬁ brous Hand augering revealed a continuous layer of On the down-thrown side of the fault, we wood from the faulted silt (unit si-9b) below peat above the graben’s down-dropped substrate were unable to excavate deeper than ~2 m this unconformity to the SE of fault 2 yielded of terrace gravel. This peat thickens southward because of wet, unstable ground, so the base of an age of 897–722 cal yr B.P. (Fig. 6C). We to at least ~3.8 m near the southern boundary the peat was not exposed in either trench. In the are uncertain whether this sample was detrital fault (Fig. 7C), where a large piece of wood NW parts of both CC-1 and CC-4, fault strands or a root fragment. If it is detrital, then its age (sample Auger-3) was intersected at ~1 m above cutting peat terminate abruptly upward beneath (897–722 cal yr B.P.) must predate the penulti- the base of the peat (Fig. 7C; Table 1). This wood younger, unfaulted sediments along a deposi-

mate earthquake at this site (Rv2). If it is a root was part of a log (others were extracted from tional contact that we refer to as the “intra- fragment, then it does not necessarily predate the peat by the digger) that yielded a 14C age of peat unconformity.” On the NE wall of trench that earthquake. Sample RVL-2 must postdate 5580–5300 cal yr B.P. In the nearby trenches, CC-1, fault strands 4 and 5 are overlapped by the penultimate earthquake. four other wood samples (CC-4-11, CC-2-37, the undeformed peat unit, pt4 (Fig. 8A). The In trench RV-1, near-surface bulging of the CC-2-35, CC-2-33 in Table 1; also Figs. 8 and 9) SW wall of CC-1 collapsed before it could be terrace sediments is inferred to have raised the yielded 14C ages that are indistinguishable (at logged, as did the NE wall of trench CC-4. main scarp intersected by that trench, at least in 95% confi dence) from this age. A fi fth wood On the latter, we photographed and sampled part during the 1855 earthquake. Six measure- sample in trench CC-2 (CC-2-38) was slightly another exposure of the intrapeat unconfor- ments of coseismic throw on the southern part younger (5209–4842 cal yr B.P.). This suite of mity just before the wall collapsed (Fig. 8C). of the Wairarapa fault’s central section from the six similar-aged wood samples was collected There, a steep fault strand (probably equivalent 1855 event range up to a maximum of 2.5 m from different levels within the peaty graben to fault strand 4 or 5 in CC-1) juxtaposes peat in a NW-up sense (Rodgers and Little, 2006). infi ll (not just at its base). The 14C ages of these against silt. This fault is depositionally over- The height of the main scarp at Riverslea, ~3 m other wood fragments are everywhere older than lapped by an unfaulted layer of silt. The intra- (Fig. 5), suggests its growth in two stages. If that of underlying peat samples. For this reason, peat unconformity was not exposed on the SW so, these must postdate sample RV-1; that is, we interpret these wood fragments to have been wall of CC-4 (Fig. 8B). the penultimate earthquake must be younger recycled from a downed forest horizon near the Discontinuous bodies of clastic sediment abut than ca. 1300 cal yr B.P. base of the peat (the one sampled by the auger), the fault zone to the SW and are interlayered The lack of a fault in trench RV-1 and the and we establish the stratigraphic chronology in with peaty units to the north. These clastic units small offset in RV-2 suggest that the primary the trenches chiefl y on the basis of 14C ages of include cw1, cw2, and cw3 in trench CC-1, and mode of deformation to create the northwest-up peat, rather than of wood. units cwA, cwB, and cwC in trench CC-4. All scarp on the late Holocene terrace was folding. but two of these units are thickest at or near the The cause of folding is interpreted to be dip slip Trenches across the Southern Bounding fault, and pinch out northwestward into peat to on a northwest-dipping fault (concealed beneath Fault of the Cross Creek Graben defi ne a wedge-shaped body that is >1.5 m long. the trenched gravels) that forms a part of the All are truncated on their SE side against a fault, (here complex) Wairarapa fault zone. Based on Stratigraphy and Structure of Trenches except unit dp, which occurs as small isolated the stratigraphic mismatches across fault strand CC-1 and CC-4 blob in peat (in CC-4), and the large faulted 2, we infer that a small (several-meter?) strike- Adjacent trenches CC-1 and CC-4 were wedge cw2 (in CC-1), some of which extends slip surface displacement accumulated in 1855 excavated across the steep, ~2-m-high scarp across the uplifted side of the fault zone. At the on a steep splay fault in the hanging wall of the defi ning the southern margin of the pull-apart NW end of cw2, the basal contact of this wedge blind structure in response to dextral-reverse graben (Figs. 8A and 8B). Enlarged versions of truncates an underlying tree in growth posi- slip on the main structure at depth. the logs for these two trenches are provided as tion. The clastic units are interpreted by us to Fig. C, and detailed unit descriptions are given be colluvial bodies derived by redeposition of The Cross Creek Pull-Apart Graben in Appendix D. The fault zone consists of an terrace gravels eroded from the uplifted side of ~1-m-wide zone of up to fi ve fault strands that the fault. We distinguish two of them (unit cw1 Four trenches were excavated across oppo- dip steeply NW. Slivers of clay-rich sheared in trench CC-1 and the aforementioned blob, dp, site sides of a pull-apart graben along the gravel and peat (unit sg) occur between some in trench CC-4) as being smaller (<40 cm long), Wairarapa fault at Cross Creek (Figs. 1A and fault strands. Fault strand 1 (in both trenches) more organic-rich, and more lenticular in shape 7A; Fig. A). Trenches CC-1 and CC-4 were is composed of an ~5-cm-thick zone of clay than the other, wedge-shaped bodies, leading excavated across the graben’s southern bound- gouge. Only fault strand 2 (in both trenches) us to interpret the small gravel-bearing bodies ing fault (Fig. B-i), whereas trenches CC-2 and appears to cut upward into the modern soil pro- differently (see following). The youngest collu- CC-3 were excavated across its northern mar- fi le. To the NW of the zone, a sequence of peats vial wedge in trench CC-1 (cw3) is unfaulted by gin (Fig. B-ii). The swampy pull-apart graben (units pt1, pt2, pt3, pt4, pt5, and peat undiff.) strand 3, but it is not in contact with fault strand is today watered by a small southward-fl owing occurs interfi ngered with locally derived units 2. This wedge is overlain by an organic silt layer stream that traverses a series of diffuse scarp- of clastic sediment, which we interpret as scarp- containing fragments of steel wire (unit wl). The lets on the north side of the main depression derived colluvium. These down-faulted basinal youngest clastic wedge in CC-4 (unit cwC) is before it exits from the western end of the gra- deposits are juxtaposed against fl uvial terrace truncated by fault strand 2 in that trench and ben (Fig. 7A). The drier eastern end of the gra- gravels to the SW of the fault zone (units tgr-1 appears to be depositionally overlain by a unit ben is abutted by a tilted terrace surface over- to tgr-4, tg, stg, and bg). The cobble-bearing of organic silt (unit mo) containing a line of lain by a small inactive alluvial fan (Fig. 7B; terrace gravels are locally interbedded with cobbles (queried contact in Fig. 8B). Fig. B-ii). The graben is down-faulted into the sandy layers (ss, bss, gs, stg, and sd). In trench Eleven samples were 14C dated from the regionally extensive, post–Last Glacial Maxi- CC-1, an ~20-cm-thick layer of peat (pt6) is NE wall of trench CC-1 (Fig. 8A), six from mum (LGM) Waiohine terrace gravels (Begg interbedded with gravel and sand at a depth of the SW wall of trench CC-4 (Fig. 8B), and and Johnston, 2000). ~2.3 m below the terrace tread. two from the unlogged (collapsed) NE wall of

14 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

stream scarplets

trenchestrenches

Cross Ck Rd N approx scale 50 m

B 2698120 2698160 3030 ed n U ifi io 27.527.5 terraceterrace U d ct terraceterrace o ru m st Wairarapa Fault surfacesurfac ts n 6001660 le co rp k U ca ac scarpletss tr modified 252524 byby track construction D 22.522.5 D small fanfan D ? CC-3 2121 CrCross Ck.C Road U oss Ck. 21.521.5 auger transect k ? CC-2 . RoadRoad 6001620 (Fig. 7e) fenceline aaugeruger 2020 holesholes 21.521.5 terraceterrace 2698080 swamp surface N CC-1 00 CC-4 19.519.5 220220

(datumn = 1505 is M.S.L.)Points 2698200 Contour interval = 0.5 m 6001580 location of road, tracks, 20 m 20 m fence are approximate only

C NW SE approx. area trenches these auger holes trenches area of of Fig. 9 CC-2, CC-3 offset to NE CC-1, CC-4 Fig. 8 gro und surface of pull-apart graben approx elev auger holes sample above M.S.L. 2 m t ? op o peat “Auger-3” 18 m Vertical = 2x f ter basin Horizontal Scale race grave 4 m ls 16 m buried fault? Figure 7. (A) Vertical aerial photograph of Cross Creek pull-apart graben showing trench sites and fault-trace locations. The photograph by Lloyd Homer (GNS Science) predates track construction that has mostly destroyed the scarplets on the NW side of the graben. (B) Microtopographical map of the NE end of the pull-apart graben showing trench locations. Map is based on 1505 points that were surveyed using real-time kinematic differential global positioning system (GPS) methods. GPS survey was tied to a base station at benchmark LUCENA (NZMS270-S27A). Grid marks (in meters) refer to the New Zealand Map Grid Coordinate System. (C) Auger transect and topographic proﬁ le across Cross Creek pull-apart graben showing depth to gravel and location of the 14C sample Auger-3. See part B for transect location. Note that scale of the proﬁ le is slightly enlarged relative to B, and that it is vertically exaggerated by 2:1. M.S.L.—mean sea level.

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 15

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

A Cross Creek Trench 1 (CC-1), NE Wall : Southern strand, Wairarapa Fault Zone

NW 6 Radiocarbon samples with field number (all samples in this trench start with the prefix, “CC-1-”) SE

“Waiohine Terrace” base of topsoil

topsoil removed buried by digger tgr-4 fence cw2 wire rp sca om fault bss n atio swamp cav h ex topsoil ss nc Fault, showing sense wl 14 re f t 1 * o of dip-separation * se 13 ba cw3cw3 tgr-3 (strand number Vertical Distance (m) Distance Vertical cw2cw2 6 7 pt5pt5 sgsg is circled) 1 m treetree rrootsoots 8 cw1cw1 1 tgr-2 (in(in ssitu)itu) peat 9 ppt4t4 Individual cobbles undiff. 10 Horizontal = 0.5 sg (where differentiated) intrapeatintrapeat 2 pt6 Vertical Scale 12 11 3 1a-i unconformityunconformity 1 m 4 5 tgr-1 1a-ii pt1 pt2 pt3 Horizontal Distance (m) 0.5 Sheared silt-matrix gravel with subangular clasts that are locally aligned parallel to fault sg (includes 6 cm of plastic clay gouge along strand no. 1) wl Organic silt with gravel clasts. Unit includes embedded metal wire and is informally termed the “wire layer” cw1 cw2 Angular- subangular gravel in silty and sandy matrix. Interpreted as colluvium. cw-1 has organic-rich matrix. cw3 bss Coarse sand, fine pebbly sand, and sandy pebble gravel (bedded at dm scale). Interpreted as fluvial.

ss Silty fine-medium sand with coarse sandy laminae. Interpreted as fluvial. ptn Units (n = 1,2,3,4,5,6) of variably silty, organic clay and peat with abundant plant fibers, wood fragments om Dark organic clay with scattered granules to fine pebbles and charcoal. tgr-n Units (n = 1, 2, 3, 4) of massive to crudely bedded, pebble-cobble gravel with sandy matrix. Interpreted as fluvial.

Figure 8. Logs of trenches CC-1 and CC-4 across the southern bounding fault of the Cross Creek pull-apart graben. See Figures A, 2, 5, and 7B for location of trenches, Figure B-i for a photograph of the trench sites, Appendix D for detailed stratigraphic descriptions, and Table 1 for 14C analytical results of the numbered samples: (A) Log of Cross Creek 1 trench (CC-1). (Continued on following page.)

CC-4 (Fig. 8C). Of these, three are wood, and to the remaining set of nearby samples (from ~5.2 k.y. An older (sixth) event occurring after 16 are peat or organic clay-silt (Table 1). Two both above and below the unconformity). After 12–13 ka is expressed in trench CC-1 by over- samples (CC-1-1a-i and CC-1-1a-ii) were col- considering the full distribution of 14C ages, we lap of fault strand 3 by terrace gravel (unit tgr- lected from a peat that occurs interbedded with interpret CC-1-11 and CC-4-3 as recording non- 3), but the minimum age of this earthquake is alluvial gravels on the uplifted fault block. depositional events, though the reasons for this poorly constrained, and this event will not be These yielded ages between 13,160 and 12,100 are unknown (see Appendix E). Our interpreta- discussed further. In Table 2, the ﬁ ve youngest

cal yr B.P. (at 95% confi dence). The rest of the tion honors the full complement of remaining events are labeled CCS1 (youngest earthquake on 14 samples were collected from the infi ll of the gra- C ages, whereas any other explanation for the the southern boundary fault) to CCS5 (fi fth-old- ben on the downthrown side of the fault, yield- age reversal would require additional ages to be est earthquake on the southern boundary fault). ing ages of <5500 cal yr B.P. One wood sample rejected. It is important to note that other inter- Table 2 also identifi es the specifi c 14C samples (CC-4-11) yielded an age that was greater than pretations of the 14C data would not affect the that constrain the maximum and minimum age that of underlying peat samples, and it is inferred total number of surface rupturing events inter- limits for each of these earthquakes. All event to have been recycled from a near-basal part of preted from the Cross Creek trenches. ages are quoted in calibrated years B.P. at the the peat sequence. All the other samples in these 95% confi dence interval. Figure 10 plots these trenches yielded ages that are in the correct Interpretation of Earthquakes Rupturing age ranges and also labels the key 14C samples stratigraphic order (at 95% confi dence), with the Southern Bounding Fault of the Graben used to bracket these age ranges. the exception of two peat samples (CC-1-11 and On the basis of the combined data from There is an along-strike difference in exposed CC-4-3) that were collected in proximity (above trenches CC-1 and CC-4, we interpret fi ve earth- stratigraphic age between the two southern and below) to the intrapeat unconformity. These quakes to have ruptured the southern bounding trenches; older sediments are exposed in CC-4 yielded ages discordant to one another and fault of the Cross Creek graben during the past compared to in CC-1. We attribute this difference

16 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

B Cross Creek Trench 4 (CC-4), SW Wall (inverted view): Southern strand, Wairarapa Fault 6 NW Radiocarbon samples with field number (all samples in this trench start with the prefix, “CC-4-”) SE “Waiohine Terrace” base of topsoil

face ground sur

cobble line ? (human-disturbed p ar wedge?) t sc faul tg swamp ? cscs sd Fault, showing sense topsoil momo cwB 1 of dip-separation gs (strand number 1313 stg 1 m cwCcwC sisi 16 cwAcwA is circled) 1 ccwBwB 11 2

Vertical Distance (m) Distance Vertical 10 6 bg Individual cobbles ccwAwA tion Horizontal = 0.5 6 va 4 ca (where differentiated) ddpp ex Vertical Scale peat undiff. of se ba sg sg 1 m Horizontal Distance (m) 0.5 mo Dark brown organic clay-silt (grades NW into peat). Includes line of cobbles. C intrapeat unconformity si, cs si is medium gray, fine pebbly silt, grading NW into peat. cs is light-gray clay-silt 1.6 m below ground cwA Wedge-shaped units (n = A, B, C) of subangular pebble gravel peat cwB supported by clay-sand matrix (interpreted as colluvium). CC-4-3 cwC siltsilt overlapoverlap dp Deformed lens of peat that includes entrained cobbles and pebbles peat, undiff. Organic clay and peat with abundant wood fragments and roots sg Soft, sheared pebble-cobble gravel CC-4-2 (includes 5 cm of clay gouge and 5 cm of smeared peat along fault strand no. 1) clay-clay- tg Terrace gravel: crudely bedded, cobble gravel with granular sandy matrix (fluvial) peat siltsilt sd Medium sand with scattered pebbles (well-sorted, fluvial)

gs Gravelly (pebbles, cobbles) medium sand stg Pebble-cobble terrace gravel, stratified 10 cm fault bg Bouldery terrace gravel

Figure 8 (continued). (B) Log of the nearby Cross Creek 4 trench (CC-4). View has been inverted for ease of comparison with Figure 8A. (C) Photograph of unlogged NE wall of trench CC-4 just prior to its collapse, showing exposure of fault-draping intrapeat unconformity and location of 14C samples CC-4-2 and CC-4-3. An enlarged version of this ﬁ gure is presented as Figure C.

in exposure level to slight up-bulging of strata over of fault strands 4 and 5 by peat in trench record death of a tree by earthquake-induced along the fault near CC-4 (see Appendix E). CC-1 (Fig. 8A) and the unconformable over- toppling immediately prior to emplacement of Because of this NE structural plunge along the lap of the unnamed fault by silt in trench CC-4 the wedge. In trench CC-4, the same earthquake

fault, only trench CC-4 recorded evidence for the (Fig. 8C). We bracket the unconformity (and (CCS3) is interpreted to have formed the wedge,

oldest event (CCS5). CCS5 resulted in the forma- thus CCS4) to the interval 3690–2970 cal. yr B.P. cwC. A maximum age constraint for wedge tion of the large colluvial wedge (cwA) exposed using samples CC-1-12 and CC-1-10. A com- cwC of 3340 cal. yr B.P. is provided by sample in the lower part of trench CC-4. Radiocarbon bination of the age constraints of the colluvial CC-4-13, which underlies it (there are no dated samples CC-4-6 and CC-4-10, from below and wedge with those of the unconformity yields a samples from above the wedge).

above this wedge, bracket this earthquake to the composite age range for event CCS4 of 3690– In trench CC-1, we infer that the penultimate

interval 5450–4620 cal. yr B.P. 3070 cal. yr B.P. earthquake (event CCS2) caused emplacement

We interpret the next-youngest earthquake, The third-youngest earthquake (event CCS3) of a colluvial wedge, cw3. This wedge draped

CCS4, to have caused refreshment and collapse rejuvenated the scarp to cause formation of col- across the preexisting (and gouge-laden) trace of the scarp, leading to emplacement of the colluvial wedge cw2, as exposed in trench CC-1. of fault 1. The age of the wedge is bracketed by luvial wedge, cwB. The age of this wedge is The age of this wedge is bracketed by samples samples CC-1-13 and CC-1-14 to the interval bracketed by samples CC-4-16 and CC-4-13 to CC-1-6 and CC-1-13 to the interval 2340–740 920–800 cal. yr B.P. We infer that the upper- the interval 4870–3070 cal. yr B.P. In addition, cal. yr B.P. Our preferred age for this earth- most units in CC-4 are condensed or have been we infer that the same earthquake was recorded quake is based on sample CC-1-6 alone, as this in part eroded, perhaps as a result of deforesta- by the intrapeat unconformity—the draping- age (2340–2110 cal yr B.P.) is interpreted to tion and agriculture. In CC-4, a diffuse line of

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 17

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

A Cross Creek Trench 2 (CC-2), NE Wall: Northern strand, Wairarapa Fault Zone 31 Radiocarbon samples with field number (all samples in this trench start with the prefix, “CC-2”)

NW fence-line SE

base Fault, showing sense scarp 1 m 37 of topsoil of dip-separation 1 Horizontal = ss (strand number Vertical Scale 0.5 co-3co-3 gsi is circled) tg 4 Co-2Cco-2o-2 road fill 1 m ss 31 0.5 sg sisi

38 Vertical Distance (m) ptpt 35 33 osi-1 osi-2 tgtg 30 ground surface (swamp) 34slt 3344 pt,pt, uundiffndiff deformational bulge 12 3 co-1 tg slt Individual cobbles

(where differentiated) of excavation base

Horizontal Distance (m) gsi Gravelly silt, very poorly sorted (interpreted as anthropogenic fill related to road construction) ss Coarse sand (moderately sorted, fluvial)

osi-2 Dark brown organic silt with crumb structure and abundant charcoal fragments throughout si Yellowish-grey to orangish-brown or v. dark gray silt and fine sand, charcoal-rich at base osi-1 Organic silt with crumb structure and rare pebbles pt Peat: dark-brown fibrous, locally silty, some pebbles. Abundant detrital wood, charcoal; includes whole logs in swamp pt undiff.

sg Sheared terrace gravel with alignment of clasts parallel to fault

co-3 Gravelly silt: gravel clasts (subangular-subrounded) in silt matrix. Interpreted as colluvial wedge

co-2 Gravelly sand: gravel clasts (subangular-subrounded) in sandy silt matrix. Interpreted as colluvial wedge

co-1 Layer of rounded cobbles. Interpreted as terrace-derived gravel (redeposited colluvium)

slt Fine sandy silt and silty fine sand with subangular pebbles (interpreted as overbank deposits)

tg Pebble-cobble terrace gravel, with some lenses of coarse sand, crudely bedded to massive (fluvial)

Figure 9. Logs of trenches CC-2 and CC-3 across the northern bounding fault of the Cross Creek pull-apart graben. See Fig- ures A, 1, 5, and 7B for location of trenches, Figure B-ii for a photograph of the trench sites, Appendix F for detailed stratigraphic descriptions, and Table 1 for 14C analytical results of the numbered samples. (A) Log of NE wall Cross Creek 2 trench (CC-2). An enlarged version of the log is presented as Figure D. (B) Log of NE wall of Cross Creek 3 (CC-3). (C) Log of SW wall of CC-3. View is inverted for ease of comparison with parts A and B. An enlarged version of the CC-3 logs are presented in Figure E.

cobbles in the mo unit near the ground surface bance caused by deforestation and cultivation applied, however, to the two smallest (tens of may be a human-disturbed equivalent to wedge must have removed any evidence for it from centimeters long) gravel-bearing bodies near cw3, but this is uncertain. the uppermost sediment layers. the scarp of the southern boundary fault: spe- Inferred on historical grounds to be the 1855 ciﬁ cally, units cw1 in CC-1 and dp in CC-4.

earthquake, the most recent earthquake (CCS1) Inferring Earthquakes from Colluvial This difference in interpretation was based on is expressed in trench CC-1 by the rupturing Wedges our consideration of the following typical attri- of fault strand 2 upward from wedge cw3 to These interpretations rely in part on our inter- butes of earthquake-induced colluvial wedges: extend into the modern soil proﬁ le. A maxi- pretation that the large (>1.5-m-long) gravel (1) large size (consistent with generation of a mum age constraint of 970 cal. yr B.P. is pro- wedges are scarp-derived colluvial units that meter-high fault scarp); (2) wedge shape (thick- vided by sample CC-1-14 from the faulted om formed as a result of fault-scarp rejuvenation est at the fault, where they may be truncated, layer, which is overlain by the wire-bearing during earthquakes. We infer that in the densely and thinning away from the fault); (3) compo- layer, wl (undated, but assumed modern). In forested lowland settings of precolonial New sition of clasts equivalent to exposures in the trench CC-4, the 1855 earthquake is inferred Zealand, gravitational collapse of earthquake- adjacent fault scarp; (4) texture consistent with to have caused slip on fault strand 2, but it did induced fault scarps was the chief process by transport of these clasts down the scarp; and not generate any colluvial wedge that is pre- which large bodies of terrace gravel could be (5) synchroneity to other wedges along the served near the ground surface today. If such eroded from a fault scarp and redeposited in an fault (or at least to other types of evidence for a wedge once did exist, human-induced distur- adjacent peat basin. This interpretation was not earthquakes along that fault).

18 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

B Cross Creek Trench 3 (CC-3), NEWall: Northern strand, Wairarapa Fault Zone 2 Radiocarbon samples with field number (all samples in this trench start with the prefix, “CC-3”)

ground surface NW base of topsoil Fault, showing sense SE ssss of dip-separation 1 sp2 (strand number ccpgpg is circled) grs s-col 1 m u-col Horizontal = sssss roadr fill pt5 tg 7 oad Vertical Scale 0.5 base of trench fill pgpg 2 cco-2o-2 1 m 1 2 0.5 swamp sand-pebble op pt4 6 fluvial marker pt4

Vertical Distance (m) gs layer fissure 3 stg tg pt2 ogs ssg Individual cobbles (where differentiated) sp1 os cs pt1 tree pt3

Horizontal Distance (m) Cross Creek Trench 3 (CC-3), SW Wall (inverted view): Northern strand, Wairarapa Fault Zone C L Radiocarbon samples with field number (all samples in this trench start with the prefix, “CC-3”) NW SE ground surface Fault, showing sense of dip-separation grsgrs base of topsoil ss piece (strand number is circled) of wire 1 m cpgcpg Horizontal = 6 1 5 Vertical Scale 0.5 ss u-col tg grsgrs roadro fill 1 m s-col ad f 0.5 base of trench co-2co-2 ill sand-pebble ssg 2 F pt swamp fluvial marker layer 1 3 E Vertical Distance (m) fissure and fissure infill gs pg stg fg L co-1 Individual cobbles (where differentiated) pt tg

Horizontal Distance (m) Explanation u-col Pebble-cobble gravel in a sand-silt matrix (v. poorly sorted).Clasts are angular-subangular. s-col s-col contains wire and both units are interpreted as anthropogenic fill grs Coarse gravelly silt to NW of fault zone, grading SW into fine to coarse gravel that contains increasing sand-silt component.

op Organic-rich silt and peat. Pronounced crumb soil structure. Abundant modern roots

cpg Pebble gravel (compact). Clasts are well-sorted, subrounded, and imbricated (interpreted as fluvial deposit)

ss Coarse sand (well-sorted) with sparse fine pebbles (fluvial)

pg Coarse sand to fine pebble gravel, locally including cobbles. Interpreted as fissure-infilling fluvial deposit

co-2 Pebble-cobble gravel supported by matrix of silty or clayey sand (poorly sorted). Clasts are subangular-subrounded. co-1 Interpreted as colluvial wedges

ogs Mixed unit near tree: Gravelly organic rich silt. Subrounded-subangular clasts cs Mixed unit near tree: Gravelly coarse sand os Mixed unit near tree: Organic rich silt with varying gravel component. Subrounded-subangular clasts

pt1,pt2 Fibrous, organic-rich clay or peat with abundant wood fragments and some scattered pebbles pt3,pt4 sp1 sp2 Tectonically mixed and sheared peats and gravels (subvertical orientation of blade-shaped clasts in sp1)

gs Gravelly sand in matrix of sandy silt, clast supported (fluvial terrace)

ssg Organic rich, fine sandy and silty gravel, subrounded-subangular clasts (fluvial terrace deposit)

stg Sheared terrace gravel with alignment of clasts parallel to fault (coarse sandy matrix) fg Faulted cobble terrace gravels tg Terrace gravel: boulders and cobbles (subrounded) in a pebble-sand matrix. Clast supported and crudely bedded.

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 19

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

Palliser Bay Uplift Wairarapa F. Wharekauhau F. Wairarapa F. Wairarapa Fault Turakirae Hd Lake Cross Creek Beach Ridges (BR) Kohangapiripiri Riverslea South: CC-1 & 4 North: CC-2 & 3 Pigeon Bush Tea Ck. Road Event Names: Tk K Rv CCs CCn Pb T

cal. yrs BP BR-2 channel K1 RVL-3 ? 0 Tk1 abandonment Rv1 ??CCs1 CCn1 1 AD 1855 6.4 m slip on fault 2 Pb T1 trench ? trench PB-5 RVL-2 PB-1 RV-2 CCs2 PB-1 Rv2 CC-1-14 CC-1-13,-14 Pb2 Penultimate 1000 RVL-4 CC-1-13 wedge cw3 CCn2 burn event (pre-abandon- decimeter-scale wedge cw2 ment) T2 organic-matrix in trench CC-1 BR-3 gravel lenses CCs3 CC-3-2 2000 Tk2 CC-2-31 K2 Third Event 9.1 m CC-1-8 CC-3-2 cw-1 CC-1-6 CC-1-6, -7 CCn3 wedge co-2 in unit in (tree death age) trench CC-3, CC-2 trench CC-1 CC-1-10 T3 3000 CC-1-9 CC-4-13 CC-3-F, -E & -6 CC-1 CC-4 CCs4 Fourth Event K3 wedge cwB CC-1-12 CC-3-E unconfor- not in trench CC-4 wedge co-1 4000 mity dated CC-4-2 in trenches CC--3 and CC-2 BR-4 (model age) 3 CC-4-10 4 Tk CC-2-30, -34 T 6.8 m CC-4-4 5000 CC-4-16, -10 CCn4 Fifth Event K4 dp lens in trench CC-4 CCs5 CC-4-6 CC-3-L CC-4-6, -11 wedge cwA detrital wood in trench CC-4 subpopulation 6000 (downed forest?) pre-CCn4 T5 BR-5 (Auger-3, CC-4-11 CC-2-38 CC-2-38, CC-2-37, Tk4 youngest wood sample 7.3 m K5 CC-2-35, CC-2-33) 7000 in recycled subpopulation = age of forest destruction?

Explanation: CC-4-10 95% age range for earthquake 8000 at individual trenching site, (composite) 95% confidence CCn4 showing bracketing sample numbers age range on earthquake and local name of event preferred age CC-4-12 of earthquake

Figure 10. Time-space plot of surface-rupturing Wairarapa fault earthquake events inferred from all available types of paleoseismological data (Tables 1, 2, and 3). Data for Tea Creek trench (events labeled, T) are from Van Dissen and Berryman (1996). Diatom-based paleoenvironmental data for coastal uplift events at Lake Kohangapiripiri (events, K) are from Cochran et al. (2007). Uplift events near Turakirae Head on Palliser Bay (events, Tk) inferred from raised beach ridges (BR-1 to BR-5) are taken from McSaveney et al. (2006). Small numbers quoted in meters denote inferred single-event uplift magnitudes for a given event. For Palliser Bay data, these refer to the maximum uplifts near Turakirae Head as inferred by McSaveney et al. (2006). Name of key 14C samples that bracket the timing of rupture events in trenches (this study) are identifi ed at maximum and minimum limits of error bars (95% confi dence) and refer to the fi eld sample names listed in Table 1. Colored horizontal bands depict the maximum and minimum age range (at 95% confi dence) for each earthquake event as derived from an analysis of the composite set of trench data and 14C results.

20 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

The larger gravel bodies that we have inter- potential ambiguity of earthquake identifi ca- cpg (upper layer of pebble gravel). These preted as earthquake-induced colluvial wedges tion. Enlarged versions of the trench logs are loosely consolidated and iron-stained deposits fi t all of these criteria, whereas the two decime- provided in Figures D and E, and detailed unit are distinctively well rounded and well-sorted, ter-sized bodies (units dp and cw-1) fi t few of descriptions are given in Appendix F. and we interpret them to be fl uvial. On all three them. The size of the former is most plausibly None of the fault strands on the northern side logged walls, the fl uvial deposits deposition- attributed to the scale of coseismic shaking, of the graben cuts to the surface, and the topo- ally overlie the terrace gravels (unit tg) on the scarp rejuvenation, and subsequent scarp ero- graphic scarp is offset ~5–6 m southward rela- NW side of fault strand 1 (Figs. 9A, 9B, and sion that would accompany inferred fault throws tive to the subsurface fault zone. These relation- 9C). Between fault strands 1 and 2, an intact of ~1–2 m on the Wairarapa fault (Rodgers and ships refl ect lateral accretion of the uppermost (unfaulted) part of the fl uvial sequence deposi- Little, 2006). By contrast, the other two bodies gravelly layers across the fault scarp by anthro- tionally overlies a substrate of strongly sheared are small, isolated (dp is a disconnected blob), pogenic processes. Wire fragments found in the gravel mixed with clay pug (unit sg in Fig. 9A; and have a lenticular rather than wedge shape. s-col unit (Fig. 9C) indicate that both this unit unit stg in Figs. 9B and 9C). In trench CC-3 Although both types of gravel body have appar- and the overlying u-col consist of fi ll material (Fig. 9B), the pebble-rich basal part to fl uvial ently been derived from redeposition of the ter- pushed southward from the site of the nearby package (pg unit) thickens abruptly downward race gravels derived from the uplifted side of the road during its excavation (Fig. 7B). to occupy the space between fault strands 1 fault, clasts in the decimeter-sized gravel bod- Fluvial terrace gravels (tg) in the uplifted foot- and 2. We interpret this to be an infi lled fault- ies are supported by an organic-rich or peaty wall of the fault zone are juxtaposed across the fi ssure. In CC-3 (NE wall, Fig. 9B), the upper matrix, whereas the matrix of the large wedges northern boundary fault zone against a basin to part of the fl uvial infi ll (cpg unit) drapes south- is organic-poor. Perhaps the small bodies formed the SE that is dominated by peat and organic silt. eastward across fault strand 2 to lie on the peat by adhesion of gravel clasts onto the roots of On the SE side of the fault, the down-dropped ter- of unit pt5. In trench CC-2 (Fig. 9A), the col- toppled trees. Excluding the 1855 earthquake race gravels form an exposed depositional sub- luvial wedge (co-3) depositionally overlies the (evidence for this youngest event appears not to strate to the peat-rich basin fi ll. The terrace grav- sandy ss layer (part of the fl uvial infi ll pack- be well-preserved today), all the “large” wedges els (tg) are locally capped by layers of gravelly age). The wedge co-3 overlaps fault strands 1, at Cross Creek can be temporally correlated to sand and silt (units slt, gs, and ssg). These fl uvial 2, and 3 and is displaced by fault strand 4. other wedges (or at least earthquakes) in one deposits are overlain by a much fi ner-grained Seven samples from trench CC-2 and fi ve or more other trenches, implying their lateral sequence of organic-rich silts and peats (units pt, from CC-3 were 14C dated (Table 1). Of these, continuity along the fault. By contrast, none of osi-1, osi-2, op, and pt1, pt2, pt3, and pt4). About four are wood, and the rest are peat, organic the two cited “small” redeposited gravel bodies, 7 m to the SE of the fault, a deformational bulge clay, or charcoal. All from trench CC-2, the although they are well dated (Fig. 10), could be is expressed by folding and erosion of part of the wood samples (CC-2-33, CC-2-35, CC-2-37, recognized beyond a single trench wall. Next, organic-rich basinal infi ll (Fig. 9A). CC-2-38) yield ages that are older than strati- we will show that most of the colluvial wedges Near the fault, the peaty basinal units are graphically underlying nonwood samples (CC- at Cross Creek can be correlated across both interfi ngered with three southward-tapering 2-30 and CC-2-34) but are indistinguishable sides of the graben. gravel wedges. These wedges are progressively from one another (ca. 5.6–5.0 ka). As explained faulted and tilted. They are labeled, from old- already, the wood samples are interpreted to have Trenches across the Northern Bounding est to youngest, co-1, co-2, and co-3 (Figs. 9A, been recycled from a near-basal forest layer that Fault of the Cross Creek Graben 9B, and 9C). Truncated against the fault zone, was disrupted soon after inception of the pres- and consisting of poorly sorted pebble and ent graben. All the nonwood samples in trenches Stratigraphy and Structure of Trenches cobble gravel in a silt matrix, they are inter- CC-2 and CC-3 yield 14C ages that are in the cor- CC-2 and CC-3 preted to be scarp-derived colluvial wedges. rect stratigraphic order (see Appendix G). Trenches CC-2 and CC-3 were excavated The oldest of these, co-1, is best exposed on across the northwestern margin of the pull- the SW wall of CC-3 (Fig. 9C). On the NE wall Interpretation of Earthquakes Rupturing apart graben. We logged the NE wall of CC-2 of the same trench (Fig. 9B), it is complexly the Northern Bounding Fault of the Graben (Fig. 9A) and both walls of CC-3 (Figs. 9B and deformed (disturbed) adjacent to a large fossil On the basis of the combined data from CC-2 9C). Many of the stratigraphic units can eas- tree (this was removed by the digger). There and CC-3, we recognize at least four earthquakes ily be correlated between these three walls, so the gravel wedge has apparently been entrained to have ruptured the northern part of the Cross we have adopted a set of (in part) common unit into the gravel-bearing mixed units os, ogs, and Creek graben since ca. 5.2 ka. These are labeled

names that reﬂ ects our correlation (Figs. 9A, 9B, cs, and also dismembered into isolated clasts. CCN4 (oldest) to CCN1 (youngest) in Table 2, and and 9C; Appendix F). This part of the Wairarapa In trench CC-2, co-1 is expressed as a stone their age constraints are plotted on Figure 10.

fault zone consists of several SE-dipping fault line of footwall-derived terrace cobbles (well The oldest earthquake, CCN4, resulted in strands. The fault numbering in Figure 9 reﬂ ects rounded) that extends southeastward above a emplacement of the colluvial wedge co-1 at our interpretation of how these strands correlate basal peat layer (unit pt) for >2 m away from the site of both CC-2 and CC-3. Temporal between the trenches. Fault slivers of strongly fault strand 3 (Fig. 9A). The younger and vari- constraints for this earthquake are provided by sheared, clay-matrix gravel (sg and stg) and ably tilted colluvial wedge co-2 is recognized 14C samples CC-3-L (from unit pt, below the peat (sp1 and sp2) contain gravel clasts that are on all three of the logged walls. wedge, Fig. 9C) and from CC-3-E, CC-2-30, rotated to a steep dip against the fault. In addition On its NW side, the co-1 wedge is faulted and CC-2-34 (from the same unit above the to their content of colluvial wedges, sediments against a distinctive sequence of sands and wedge). These ages bracket the earthquake to along the northern margin of the graben reveal pebble gravels. Elements of this sequence are the period 5280–4640 cal. yr B.P. This interval

clear evidence of progressive deformation, such found on all walls of both northern trenches overlaps with CCS5 on the opposite side of the as differential tilting, angular unconformities, as combinations of the units pg (basal pebble graben, an event that is similarly recorded by and ﬁ ssuring. These relationships reduce the gravel), ss (a thin marker layer of sand), and a colluvial wedge (unit cwA in Fig. 8B). We

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 21

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

therefore infer that CCN4 and CCS5 represent age constraint for CCN2 of 2150–1940 cal. yr seismic trenches, and by dating and correlat- the same earthquake. Our preferred age for B.P. is provided by sample CC-3-2 in unit pt5, ing these using the 40 new 14C samples, we

event CCN4, 5209–4842 cal. yr B.P., is based which stratigraphically underlies the cpg mem- interpret a composite surface-rupturing history on the interpretation that the six similar-aged ber of the fi ssure-infi lling sequence (Fig. 9B). that has included at least fi ve earthquakes on

wood samples were derived from a forest that The fi nal rupture at the site, CCN1, caused the southern part of the Wairarapa fault since was toppled or damaged by this earthquake. renewed slip on fault strands 1 and 2 and initiation ca. 5.2 ka (Table 3; Fig. 10). This history and For this preferred age, we use the age of sam- of fault strands 5, 6, and 7 (Figs. 9B and 9C). This our intertrench correlations are summarized in ple CC-2-38, the youngest element of the death faulting deformed the fi ssure-infi lling sequence Appendix H. Although any correlation involves assemblage that we infer to have been earth- of stream sediments (units ss, pg, and cpg) that interpretation, our chronology invokes the mini-

quake-triggered. were deposited after earthquake CCN2. In addi- mum number of earthquakes allowed by the We interpret the next-youngest earthquake to tion, faulting on strand 4 deformed the colluvial data. While the dating precision for each event

rupture the northern side of the graben (CCN3) to wedge co-3 that we attribute to CCN2 (Fig. 9A). at each site differs, and the preferred age of spe-

have caused emplacement of the colluvial wedge Inferred to be the 1855 earthquake, event CCN1 cifi c earthquakes might vary according to inter- co-2 at the site of both CC-2 and CC-3. Age con- may also have resulted in deposition of the collu- pretation, we believe that the overall number of straints for this wedge are provided by samples vial layer grs at the site of trench CC-3; however, surface-rupturing events (fi ve since ca. 5.2 ka) is CC-3-F, CC-2-34, and CC-2-30 from the pt unit it is uncertain whether this is a natural deposit. a robust outcome of the data. Not unexpectedly, below the wedge (Figs. 9B and 9C) and by sam- Note that the described progressive deforma- preservation of the individual surface-rupturing ples CC-3-2 and CC-2-31 from above the wedge tion sequence (opening of fi ssure, fi lling of fi s- events in this chronology is unequal between (the pt unit in Fig. 9B; the osi-2 unit in Fig. 9A). sure, renewed faulting) requires charcoal sam- different sites and trenches because of differ- These data yield an age range for the earthquake ple CC-3-2 (2150–1940 cal. yr B.P., from pt5 in ences in exposed stratigraphic age and degree of of 3080–1991 cal. yr B.P., an interval that over- Fig. 9B) to predate two earthquakes. Conceiv- disruption or burial of near-surface layers as a

laps with CCS3 on the opposite side of the graben, ably, it might predate three earthquakes if the result of human activity. Only the Cross Creek as expressed by colluvial wedges cw2 (in CC-1, fi ssure-infi lling signifi cantly predated deposi- graben trenches sampled the oldest three events, Fig. 8A) and cwC (in CC-4, Fig. 8B). tion of the colluvial wedge co-3 in trench CC-2. whereas the Pigeon Bush and Riverslea trenches

The penultimate earthquake (CCN2) to rup- This is because this wedge stratigraphically provided timing information relating to the last ture the northern boundary fault of the graben overlies the fi ssure-infi lling ss unit (Fig. 9A). two events only (1855 and penultimate). As a caused opening of an ~1-m-deep fault fi ssure. We view such a three-event scenario as less result of such preservational differences, no sin- This cavity was infi lled initially by the pebble likely than a two-event scenario, since no other gle trench preserves evidence for all fi ve events gravel unit pg and later by the sand of the ss unit. trench provides evidence for three earthquakes (though combinations of four are recorded in These well-sorted fl uvial units were deposited since ca. 2 ka. In our preferred interpretation, each of the four Cross Creek trenches). By cor- across the pug-lined faults bounding the fi s- the fi ssuring and emplacement of the colluvial relating events between the various trenches and sure (fault strands 1 and 2 in Fig. 9B), with the wedge co-3 are viewed as twin manifestations trench walls, and using the entire data set of 14C

youngest part of the infi ll sequence (unit cpg) of the penultimate earthquake, CCN2. samples to bracket the composite event timing, downlapping onto the peat unit pt4 (Fig. 9B). we were able to signifi cantly narrow the 95% Presumably a small stream draining across the DISCUSSION confi dence time intervals for each event (red

scarp transported the clasts. We infer that CCN2 bars in Fig. 10) relative to that which would caused, moreover, some combination of the fol- Late Holocene Rupturing History of the have been derived by limiting our analysis to lowing: SE-ward tilting of the co-2 wedge, anti- Southern Wairarapa Fault 14C data found in each individual trench or site clinal bulging of the peat basin to the SE, and (compare Table 2 to Table 3). deposition of the colluvial wedge co-3 on the By integrating the key stratigraphic and The Cross Creek pull-apart graben is the key NE wall of trench CC-2 (Fig. 9A). A maximum structural events observed in the eight paleo- locality; it provides an organic-rich stratigraphic

TABLE 3. COMPOSITE AGE RANGES AND INTERVALS FOR SURFACE-RUPTURING EVENTS ON SOUTHERN WAIRARAPA FAULT BASED ON ALL OF THE TRENCHES (THIS STUDY) Rupture event Trench sites & 14C-derived maximum age range Preferred age† event numbers (cal. yr B.P., 95% confidence)* (cal. yr B.P., 95% Younger limit (sample no.) Older limit (sample no.) confidence) Most recent Rv1, Pb1 264–0 (RVL-3) 547–497 (RVL-2) 95 (=1855 AD) event Penultimate Rv2, CCs2, CCn2, 970–800 (CC-1-14) 920–740 (CC-1-13) 920–800 event Pb2 Third event CCs3, CCn3 2294–1991 (CC-2-31) 2340–2110 (CC-1-6) 2340–2110 (CC-1-6) Fourth event CCs4 3340–3070 (CC-4-13) 3690–3460 (CC-1-12) 3690–3070 Fifth event CCs5, CCn4 4980–4640 (CC-2-30) 5280–4850 (CC-3-L) 5209–4842 (CC-2-38) *Limiting radiocarbon ages. Field sample number in brackets (see Table 1). Bold type indicates extreme bracketing age (at 95% confidence). †For most recent event, inferred on the basis of historical data; for third event, assumed to be coincident with death of the toppled, in situ tree in trench CC-1, the outer layer of which was dated in sample CC-1-6; for the fifth event, assumed to be coincident with detrital wood sample CC-2-38, which is the youngest dated element of an upward-recycled subpopulation of wood ages inferred to be derived from a major local deforestation event triggered by earthquake shaking.

22 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

record that was progressively deformed near ment (e.g., Berryman, 1987; Nelson and Manley, to include both our penultimate and third events. the fault, especially on the northern side of the 1992; Wilson et al., 2007b). Turakirae Head is Accordingly, we infer that a distinct beach ridge graben. The closely spaced pair of trenches, one of the world’s best examples of a coseismi- may not have been introduced (or preserved) CC-1 and CC-4, across the southeastern margin cally uplifted fl ight of Holocene beach ridges into the landscape after the penultimate earth- of the Cross Creek pull-apart graben recorded (Burbank and Anderson, 2001). Others include quake. The vertical interval between beach all fi ve of the earthquake events. Four of these Mocha and Santa Maria Islands in Chile (Nelson ridges BR-3 and BR-4 records not only the fi fth events are corroborated in the two trenches on and Manley, 1992; Bookhagen et al., 2006), the event (in time) but also our “missing” event four. the northwestern margin of the graben (CC-2 Boso Peninsula and other sites in Japan (Ota and At 5.5 m, the height difference between these and CC-3). This correspondence implies that Yamaguchi, 2004), Peninsula de Nicoya, Costa two ridges implies a mean tectonic throw of both strands of the graben ruptured together Rica (Marshall and Anderson, 1995), the Mahia ~2.6 m for those two earthquakes (McSaveney during these four earthquakes, and it reinforces Peninsula and Pakarae River mouth in New et al., 2006). Perhaps, one of these two incre- our confi dence in the composite chronology. Zealand (Berryman, 1993; Ota and Yamaguchi, mental uplifts was too small to preserve a dis- The second-oldest earthquake in our chronol- 2004; Wilson et al., 2007a), Cape Mendocino, tinct beach ridge at Turakirae Head.

ogy (event CCS4, recorded in part by the fault- California (Merritts, 1996), Taiwan (Yamaguchi Our proposed identifi cation of the two “non– draping intrapeat unconformity) apparently and Ota, 2004), and the Gulf of Alaska (Plafker, beach-ridge” earthquakes is independently sup- ruptured only the southern margin of the gra- 1969; Plafker et al., 1992; Plafker and Rubin, ported by other studies. Using the diatom record ben, and it is the one exception to this mutual 1978). Although other studies have compared cored from Lake Kohangapiripiri (Fig. 1), rupturing of bounding faults. Subjective ele- beach ridge uplifts to the timing of historic earth- Cochran et al. (2007) inferred a sudden shal- ments of our composite interpretation include: quakes as much as ~500 yr ago (e.g., Bookhagen lowing to have taken place across that estuarine (1) attributing local 14C age reversals (relative et al., 2006; Ferranti et al., 2007), these studies lagoon at 3900–3300 cal yr B.P (this transition to stratigraphic ordering) associated with six typically suffer from the short time span of the is labeled K3 in Fig. 10). The 14C-based time similar-aged wood samples to refl ect sedimen- historic data and from ambiguities regarding the interval for this inferred uplift event overlaps tary recycling of a single forest death assem- location of the surface rupture accompanying with the trench-based timing of our fourth event blage; and (2) accepting four 14C ages near the those felt earthquakes. To our knowledge, ours (Fig. 10). Other diatom assemblage transitions fault-draping unconformity as recording peat is the fi rst study attempting a one-to-one com- reinforce the third and fi fth earthquake events deposition (these occur in correct stratigraphic parison between uplifted strandlines and paleo- of our trench-based chronology. Because the order to each other and to other surrounding seismically documented fault ruptures over a record at the top of the Lake Kohangapiripiri samples), while interpreting two other samples time span of ~5 k.y. This comparison is possible core (<2 ka) is poorly preserved, probably in part as recording nondepositional events (these because the Wairarapa fault (unlike subduction eroded, and poorly dated, it cannot be used to yield a reversed age sequence). megathrusts) is exposed above sea level where it corroborate or refute our proposed timing for the The apparent internal consistency in tim- can be investigated by trenching at sites proximal penultimate earthquake on the Wairarapa fault ing of these fi ve rupturing events between to the uplifted beach ridges on its hanging wall. (Cochran et al., 2007). Farther west on Rongotai the disparate trench sites suggests that these McSaveney et al. (2006) identifi ed and dated Isthmus (Fig. 1), Pillans and Huber (1995) broke the entirety of the southern section of the the uplift and stranding of four late Holocene identifi ed and dated beach deposits that were Wairarapa fault (Fig. 1A). By contrast, only the beach ridges at Turakirae Head. Three of these stranded above sea level at 3410–2740 cal. yr youngest (1855) and oldest of the fi ve south- are younger than ca. 5.2 ka. Each of these corre- B.P. (which correlate with our fourth event) and ern events can be correlated to the earthquake sponds to one of our independently determined at 940–260 cal. yr B.P. (which correlate with our chronology at Tea Creek trench, ~40 km to the Wairarapa fault rupturing events (Fig. 10). They penultimate event). The authors infer that most north of Pigeon Bush (Van Dissen and Berry- are the most recent earthquake (corresponds to of the shoreline uplifts on the isthmus were man, 1996) (Figs. 1A and 10). Although the uplift of beach ridge BR-2 in 1855), the third the result of Wairarapa fault paleoearthquakes; Tea Creek record is based on only one trench, event (corresponds to uplift of beach ridge BR-3 however, this inference seems uncertain given this apparently imperfect correlation suggests at 2380–2060 cal. yr B.P.), and the fi fth event the proximity of the site to the Wellington fault. that some of the southern ruptures (not includ- (corresponds to uplift of BR-4, inferred by Our comparison between the trench-based ing 1855) may not co-rupture across the north- McSaveney et al. [2006] to have taken place at earthquake rupturing history of the Wairarapa ern zone of splay fault junctions to break the 5420 –4110 cal. yr B.P.). fault and the sequence of raised beaches at Tura- entire length of the Wairarapa fault. Two of our trench-determined fault-rupturing kirae Head leads us to conclude that fl ights of events cannot be matched to a beach ridge at uplifted gravel beach ridges may provide an Comparison of Earthquake Chronology Turakirae Head. These are the penultimate event incomplete record of paleoearthquakes on adja- with that of Turakirae Head Beach Ridges and fourth event (Table 3; Fig. 10). At the crest cent reverse-oblique faults (Fig. 11A). We note For a fl ight of uplifted strandlines to provide of the Rimutaka anticline, the elevation dif- that a similar discrepancy between (less fre- a complete record of paleoearthquakes, each ference measured by McSaveney et al. (2006) quent) beach-ridge uplifts and (more frequent) earthquake must cause enough uplift to preserve between the stranded 1855 beach ridge (BR-2) earthquake events has been observed in Chile a distinct coastal landform. Moreover, these land- and the next-highest stranded ridge (BR-3) is and Italy on the basis of historical records of forms must be distinct from others that might ~9.1 m (Fig. 11A). We note that this difference earthquakes (e.g., Bookhagen et al., 2006; Cis- form by nontectonic relative sea-level changes is >2σ higher than the mean vertical spacing of ternas et al., 2005; Lomnitz, 2004; Ferranti et or longer-term tectonic processes that are aseis- the remaining three higher ridges at that site (4.8 al., 2007). The next section asks the question, mic. Finally, any interseismic subsidence due to ± 1.6 m, 1σ; McSaveney et al., 2006), implying “What processes might have led to a shortfall pre-earthquake strain accumulation or postseis- that this large vertical interval may record the in the number of beach ridge uplifts relative to mic crustal relaxation must be small or predict- cumulative uplift of two earthquakes. Tempo- earthquake ruptures on the hanging wall of the able fractions of the overall strandline displace- rally, this beach ridge interval spans from 1855 Wairarapa fault?”

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 23

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

Nontectonic Causes for Under- Representation of Earthquakes by Beach Ridges Eastern profies A Western profiles height difference between Consisting of coarse gravel, the raised 30 adjacent ridge crests beach ridges at Turakirae Head mantle a low- This interval includes BR-5 angle Holocene wave-cut platform incised into 25 3.0 m an overwhelmed or BR-4 graywacke bedrock. This is a wave-battered indistinct 5.5 m 20 coast, open to the south and the largest storm beach ridge ? BR-3 waves in Cook Strait. Since 1855, storm waves 15 9.1 m have built a berm crest on the modern beach 10 that is variably 2–7 m above mean sea level 1855 (McSaveney et al., 2006), a relationship that 6.0 m BR-2 5 BR-1 underscores the sensitivity of berm height modern to local wave conditions. Due to Turakirae Height (m) Height M.S.L.) current to (relative Head’s preeminent position on the coast, and its profile at profile at remoteness from large rivers, its rocky coastline Turakirae Head crest of anticline Horizontal Profile Distance (m) is only thinly covered with sediment. With little apparent input from longshore drift, a process (perpendicular to coast) which tends to sweep material southward away 500 m from the headland, the main source of the gravel for this sediment-starved headland has been in situ erosion of the bedrock platform (Well- man, 1967) and/or mass wasting of the adjacent B fault array at southern end hills (Hinton and McSaveney, 2007). Because of Wairarapa fault the eustatic position of sea level is inferred to have remained stable throughout New Zealand since ca. 6.5 ka (Gibb, 1986), we infer that the only signiﬁ cant source of relative sea-level fall near Turakirae Head since ca. 5.2 ka has been tectonic uplift. In a recent review paper, Ken- nedy (2008) argues that Gibb’s (1986) sea-level curve is supported by subsequent studies, and that there is no evidence for eustatic sea-level variations in New Zealand of more than ±1 m Rimutaka anticline (uplifted beach ridges) Depth

Unnamed Strike-slip faults Unnamed Strike-slip since ca. 6 ka. For this reason, we assume that Surface above folding unnamed blind (?) fault Wellington F. Wellington Ngapotiki-Ewe F. (inactive) Ngapotiki-Ewe F. Palliser F. Palliser Wairarapa-Muka Muka F. Wairarapa-Muka Wanganui Basin Ohariu F. sealevel sea-level variations have not been a key factor inﬂ uencing beach-ridge formation at Turakirae x x Head since ca. 5.2 ka. For example, we discount x AUSTRALIAN x the possibility that a small rise in sea level dur- x rface ing the Holocene (by itself) could have caused PLATE uction Inte of Subd erosion of a preexisting beach ridge. d Part ocke 40 km Given the complexity of hydrodynamic ly L inactive rent variables involved in the formation and retreat Cur inactive Wharepapa fault E of gravel beaches (Carter and Orford, 1993; AT Wharekauhau thrust Orford et al., 1995; Neal et al., 2003; Engels and PL FIC Roberts, 2005), it is possible that variations in CI PA sediment supply or wave climate (especially as 80 km the result of large winter storms) may have con- tributed to formation or preservation of discrete X’ beach ridges during some interseismic periods X of the Wairarapa fault and not others. Figure 11. (A) Beach-ridge topographic proﬁ les drawn orthogonal to the Cape Palliser–Turakirae Head The most plausible nontectonic scenarios for coastline on either side of the crest of the Rimutaka anticline (from McSaveney et al., 2006). (B) earthquake under-representation by the beach Schematic cross-section X–X′ across southern North Island, New Zealand, showing major faults in ridges probably involve some combination of upper plate of the Hikurangi subduction margin, including strike-slip and contractional faults at the the following two processes: (1) the crest of southern end of the Wairarapa fault zone. Currently inactive faults are dashed. Position of the currently locked part of the subduction interface is taken from Wallace et al.’s (2004) modeling of global the active berm may, at times, retreat landward positioning system (GPS) data. For location of cross section, see Figure 1. M.S.L.—mean sea level. far enough to weld with, or overwash, the next- highest beach ridge; or (2) a beach ridge may have been so small or indistinct at the time of its (tectonic) stranding that it was later easily

24 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

overwashed, or otherwise not preserved as a one cannot assume that there was only one rup- by sediment underplating or subduction of discrete landform on the gravelly wave-cut ture scenario during the Holocene. anomalously thick oceanic lithosphere at the platform. Secular wave climate and sediment Some earthquakes may have caused a large Hikurangi subduction zone (e.g., Litchfi eld et supply are the chief variables that control the enough coastal uplift at Turakirae Head to ele- al., 2007). Any such uplift is assumed to have size and height of a storm berm and determine vate the former storm beach beyond the reach accumulated slowly and steadily enough dur- whether it remains stable in one place (perhaps of subsequent waves, thus preserving a discrete ing the past ~5 k.y. to have had no signifi cant building up an especially high berm) or is over- beach ridge, whereas others may have not. If effect on the punctuated uplift and stranding washed to retreat landward (Carter and Orford, earthquake ruptures near the coast typically of beach ridges at Turakirae Head as a result 1993; Orford et al., 1995). If these quantities reoccupy the same spatial fault plane (or near- of large Wairarapa fault earthquakes. A point varied between interseismic periods (i.e., at a surface splay) during each earthquake, the mag- less easily dismissed is the possible effect of time scale of ~1000–2000 yr), then preservation nitude of the coseismic uplift may vary between interseismic subsidence due to elastic strain of a discrete beach ridge may have occurred earthquakes, and some may not uplift the coast accumulation above the underlying, locked after some earthquakes but not others. Given enough to strand a distinct beach ridge. Dislo- subduction interface. Permanent GPS obser- the extreme (nonlinear) impact of the lowest cation modeling of vertical deformation dur- vations along the SE coast of the North Island frequency–highest magnitude storm events ing 1855 (Darby and Beanland, 1992; Beaven (and forward dislocation modeling of GPS on the extent to which gravel beach ridges and Darby, 2005) and the extremely high slip/ data) suggest that it could be on the order of can be driven landward, a period of increased length ratio of that earthquake’s coseismic rup- 1–2 mm/yr near Turakirae Head (L. Wallace, storminess could cause an active storm berm ture (Rodgers and Little, 2006) suggest that it March 2008, personal commun.). If this rate to overtop (or weld with) a relict beach ridge may have ruptured not only the Wairarapa fault of subsidence persisted for an entire Waira- that was originally several meters (perhaps but also a contiguous segment of the subduction rapa fault interseismic period of 1–2 k.y., then even 5–8 m) higher than it; in this case, the key interface downdip of it. Perhaps some Waira- up to ~2 m of a previous coseismic uplift sig- variable would be the return period of storms rapa fault earthquakes involve deep co-ruptur- nal might be removed. However, because the large enough to generate the runup capable of ing of the subduction interface (as in 1855) to subduction interface’s earthquake recurrence overwashing the higher ridge (Orford et al., cause 3–9 m of coastal uplift locally near Tura- interval is probably much shorter (~300– 1995; Neal et al., 2002, 2003). Scenario 2 might kirae Head, whereas others break only the upper 625 yr; Wallace et al., 2004) than the Waira- occur because of either a reduced sediment sup- plate, or nucleate as strike-slip ruptures in the rapa fault’s earthquake recurrence interval ply (beach ridges cannot form without gravel) upper plate, yielding less throw at the coast. (1–2 k.y.), less subsidence is likely to accu- or because of too large a sediment supply. Alternatively, earthquake ruptures at the mulate before a Wairarapa fault earthquake The latter situation might suppress formation southern end of the Wairarapa fault may have (this amount would depend on the phase of discrete storm berms or lead to their burial a nearly constant slip and focal mechanism at shift in the seismic cycles between these two underneath younger deposits (Engels and Rob- depth, but they may cause variable coseismic faults). Either way, interseismic strain accu- erts, 2005). Clearly, the internal structure and uplift at the surface because of variable parti- mulation is predicted to reduce the height of sedimentological characteristics of the raised tioning of slip between multiple fault splays in any previously stranded beach ridge and thus beach berms at Turakirae Head might provide a the upper crust. Dislocation models by Beavan increase the possibility that it would be sub- record of such depositional or erosional events; and Darby (2005) require a local thrust struc- sequently overwashed during an especially this would seem to be a fruitful avenue of future ture to have been responsible for the very high stormy period (thus omitting one beach ridge research that could build upon (and test) the magnitude (up to 6.4 m) but short wavelength from the preserved sequence; e.g., Marshall results of our study. of uplift at Turakirae Head in 1855. Coseismic and Anderson, 1995). This would be most slip on some oblique-slip splays (especially the likely if the Wairarapa fault earthquake hap- Tectonic Causes for Under-Representation Muka Muka fault; Fig. 2B) may cause signifi - pened to occur near the end of the subduction of Earthquakes by Beach Ridges cant uplift of Turakirae Head on their hanging zone fault’s seismic cycle, and if that cycle Alternatively (and perhaps more likely), the wall, whereas slip on other splays may not. was especially long. McSaveney et al. (2006) incomplete representation of earthquakes by Consistent with this view, Schermer and Little argued, however, on the basis of a comparison beach ridges could refl ect variable magnitudes (2006) and Little et al. (2008) document a tem- between survey data and historic observations of coseismic uplift at the coast during Wairarapa porally and spatially complex pattern of fault of coseismic uplift in 1855 that interseismic fault earthquakes. Figure 11B illustrates the activation-deactivation and surface folding subsidence on Palliser Bay has been negli- complex structure of the Wairarapa fault zone at during the past 80 k.y. at the southern end of gible during the past the 150 yr. the southern end of the North Island near Tura- the Wairarapa fault zone and infer that a blind Our preferred interpretation is that coseismic kirae Head, where there are multiple (alternate) thrust is currently active near the western edge uplift has been variable at the coast near the fault strands in the near-surface. If the forma- of Lake Onoke, ~10 km to the east of the Muka Wairarapa fault, and that this tectonic variabil- tion of a beach ridge on the hanging wall of the Muka fault (Figs. 2 and 11B). These relation- ity, perhaps combined with contributing varia- Wairarapa fault zone during the late Holocene ships suggest a variable linkage between slip on tions in secular wave climate or sediment sup- was mostly controlled by the magnitude of the the deeper Wairarapa fault and its splays in the ply (or interseismic subsidence), has resulted in coseismic vertical displacement, then this would near-surface. If the 1855 earthquake involved an incomplete representation of earthquakes by have refl ected not only the direction and mag- uplift of Turakirae Head along the Muka Muka beach ridges near Turakirae Head. nitude of slip on any rupture at depth, but also fault, previous earthquakes may have ruptured the geometry of the rupturing near-surface fault to the surface along other splays to the east or Rates of Slip and Earthquake Recurrence (especially its proximity to Turakirae Head and west, causing little or no coastal uplift there. on the Wairarapa Fault its dip). Given the observed structural complex- Until now, we have ignored the possible The dating undertaken as a part of this ity of the Wairarapa fault zone near the coast, contribution of aseismic regional uplift caused study establishes the timing of deposition and

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 25

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

abandonment of the gravels comprising the 14C-derived, 10–12 ka abandonment age of the ACKNOWLEDGMENTS youngest post–Last Glacial aggradation (fi ll) Waiohine surface (~11 ± 3 mm/yr). terrace in the southern Wairarapa Valley, the We thank Kate Wilson and Vasso Mouslopou- so-called Waiohine terrace. Our 14C age of CONCLUSIONS lou for their global positioning system (GPS) samples CC-1-1a-I and Auger-3 bracket for- surveying efforts; Julia Bull, Dave Murphy, mation of the terrace tread to postdate 12,400 The Wairarapa fault is remarkable because Susanne Grigull, Vasso Mouslopoulou, and Kate ± 300 and to predate 5440 ± 140 cal. yr B.P. a series of uplifted beach ridges on its hang- Wilson for assistance with trench logging; and (Table 1). Since the ca. 12.4 ka peat sample ing wall are preserved on the nearby coast, H. Saywell, H. Brandon, P. Smith, and D. Cleal was deposited during the fi nal phases of ter- and the ages of these Holocene ridges can be for permission to excavate on their land. This race aggradation, we infer that abandonment directly compared with the ~5 k.y. chronology study was funded by the “It’s Our Fault” Proj- of that terrace took place at ca. 12 ka, making of surface-rupturing earthquakes derived from ect, with additional support provided by the New the Waiohine fl uvial terrace age-equivalent to paleoseismological methods. We dated fi ve Zealand Foundation for Research Science and either the Ohakea 2 or Ohakea 3 terrace in the surface-rupturing earthquakes on the southern Technology, contract CO5X0402 (Geo-Hazards regional correlation scheme applied to North part of Wairarapa fault since ca. 5.2 ka (an inter- and Society, GNS Science). Two anonymous Island, New Zealand (e.g., Litchfi eld and Ber- event recurrence of 1230 ± 190 yr). Along the reviewers and Jon Pelletier are thanked for their ryman, 2005). We note that our 14C-based age margins of the Cross Creek pull-apart graben, constructive comments. Helpful feedback on an data are accordant with several other ages these earthquakes caused laterally and tempo- early version of the manuscript was provided by obtained previously from samples collected rally variable amounts of faulting, tilting, and David Kennedy and Mauri McSaveney. elsewhere from above or below the Waiohine fi ssuring of scarp-proximal strata, emplacement terrace tread (Tompkins, 1987; Marden and of scarp-derived colluvial wedges, and the top- REFERENCES CITED Neal, 1990). They are also considerably more pling and destruction of trees. Our late Holo- precise than eight late Last Glacial OSL ages cene paleoearthquake record recognizes two Aston, B.C., 1912, The raised beaches of Cape Turakirae: Trans- actions and Proceedings of the New Zealand Institute, of silts that overlie the gravel of Waiohine ter- more events than are recorded by beach ridges v. 44, p. 208–213. race along the central part of the Wairarapa that were stranded on the uplifted hanging wall Atwater, B.F., 1987, Evidence for great Holocene earthquakes fault farther north (Wang and Grapes, 2007). of the fault near Turakirae Head during the same along the outer coast of Washington State: Science, v. 236, p. 942–944, doi: 10.1126/science.236.4804.942. Those samples gave ages of 16–10 ka, with time interval (McSaveney et al., 2006). Barnes, P.M., 2005, The southern end of the Wairarapa fault, four of eight ages being younger than 13 ka. Our work indicates that coseismically uplifted and surrounding structures in Cook Strait, in Langridge, Because these cover bed silts may have accu- gravel beach ridges on an open, rocky coast may R., Townend, J., and Jones, A., eds., Proceedings Vol- ume: The 1855 Wairarapa Fault Symposium: Te Papa mulated after incision of the terrace by the provide an incomplete record of paleoearthquakes Tongarewa, Museum of New Zealand. Waiohine River, their OSL ages may underes- on adjacent reverse-oblique faults. In the case of Barnes, P.M., and Audru, J.-C., 1999, Quaternary faulting in the offshore Flaxbourne and Wairarapa Basins, south- timate the timing of terrace abandonment. the southernmost part of the Wairarapa fault zone, ern Cook Strait, New Zealand: New Zealand Journal of Applying a mean OSL age of 10−11 ka to variable increments of coseismic uplift at Tura- Geology and Geophysics, v. 42, p. 349–367. displaced channels on the Waiohine terrace kirae Head may have resulted from the partition- Barnes, P.M., and Mercier de Lepinay, B., 1997, Rates and mechanics of rapid frontal accretion along the very at Waiohine River, to which they attributed a ing of slip between several different fault splays obliquely convergent southern Hikurangi margin, dextral slip of 125 ± 5 m, Wang and Grapes in the near-surface. During event four, coseismic New Zealand: Journal of Geophysical Research, v. 102, (2007) proposed a late Quaternary dextral-slip throw at the coast is inferred to have been so no. B11, p. 24,931–24,952, doi: 10.1029/97JB01384. Barnes, P.M., Mercier de Lepinay, B., Collot, J.-Y., Delteil, J., rate for the Wairarapa fault of 11.5 ± 0.5 mm/ small (<3 m) that a discrete new beach ridge did and Audru, J.-C., 1998, Strain partitioning in the transi- yr. We note, however, that these authors did not form. Variations in wave climate, sediment tion area between oblique subduction and continental not pre sent any surveyed map of these offsets, supply, and/or interseismic subsidence prob- collision, Hikurangi margin, New Zealand: Tectonics, v. 17, p. 534–557, doi: 10.1029/98TC00974. and that Lensen and Vella (1971) surveyed ably exerted further controls on the number of Barnes, P.M., Nicol, A., and Harrison, T., 2002, Late Cenozoic and mapped the same terrace in detail with- beach ridges preserved in the uplifted sequence, evolution and earthquake potential of an active listric thrust complex above the Hikurangi subduction zone, out recognizing these channels (their largest in particular, governing the morphology (e.g., New Zealand: Geological Society of America Bulletin, offset was for the terrace riser incised below height, elevation) of the original storm berm, v. 114, p. 1379–1405, doi: 10.1130/0016-7606(2002)114 the Waiohine surface, a younger landform for and whether or not it would retreat far enough <1379:LCEAEP>2.0.CO;2. Beanland, S., 1995, The North Island Dextral Fault Belt [Ph.D. which they measured a dextral slip of ~99 m). landward to overwhelm the next-highest beach thesis]: Wellington, Victoria University of Wellington, Our 14C-based dating suggests that the Waioh- ridge. A combination of such processes may have 341 p. ine terrace was abandoned near Cross Creek at caused event four to be under-represented by a Beanland, S., and Haines, J., 1998, A kinematic model of active deformation in the North Island of New Zealand, no later than 12 ka. Assuming that this surface discrete, long-lived beach ridge, despite signifi - determined from geological strain rates: New Zealand was subsequently displaced laterally by at least cant coseismic uplift during that earthquake. Journal of Geology and Geophysics, v. 41, p. 311–323. 99 m (at least near Waiohine River), our data Our 14C data indicate that a widespread Last Beavan, J., and Darby, D., 2005, Fault slip in the 1855 Waira- rapa earthquake based on new and reassessed vertical imply a minimum late Quaternary slip rate on Glacial Maximum aggradational terrace found motion observations: Did slip occur on the subduction the Wairarapa fault of ~8.3 mm/yr. On a shorter in this part of North Island, New Zealand interface?, in Langridge, R., Townend, J., and Jones, A., eds., Proceedings Volume: The 1855 Wairarapa Fault time scale, if we divide estimates of mean sin- (Waiohine terrace), was abandoned soon after Symposium: Te Papa Tongarewa, Museum of New Zea- gle-event strike slip on the southern Wairarapa 12.4 ka. This result, combined with previous land, p. 31–41. fault during the last two earthquakes (13−16 m; estimates of dextral slip relative to this terrace, Begg, J.G., and Johnston, M.R., 2000, Geology of the Welling- ton Area, New Zealand (Geological Map and Accompa- from Rodgers and Little, 2006) by our revised our revised recurrence interval for the Waira- nying Booklet): Institute of Geological and Nuclear Sci- mean earthquake recurrence interval for that rapa fault, and estimates of mean single-event ences Geological Map 10, scale 1:250,000. fault (1230 ± 190 yr), we get a dextral-slip slip on the fault, suggests that the southern part Begg, J.G., and Mazengarb, C., 1996, Geology of the Welling- ton Area (Geological Map and Accompanying Booklet): rate of 9–15 mm/yr. This result is within error of the Wairarapa fault has a late Quaternary Institute of Geological and Nuclear Sciences Geological of the above slip-rate estimate based on our dextral-slip rate of ~11 ± 3 mm/yr. Map 22, scale 1:50,000.

26 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

Begg, J.G., and McSaveney, M.J., 2005, Wairarapa fault IGCP-Project 200, in Reilly, W.L., and Harford, B.E., eds., land: New Zealand Journal of Geology and Geophysics, rupture—vertical deformation in 1855 and a history of Recent Crustal Movements of the Pacifi c Region: Wel- v. 33, p. 449–453. similar events from Turakirae Head, in Langridge, R., lington, New Zealand, Royal Society of New Zealand Marshall, J.S., and Anderson, R.S., 1995, Quaternary uplift Townend, J., and Jones, A., eds., Proceedings Volume: Bulletin, p. 377–395. and seismic cycle deformation, Peninsula de Nicoya, The 1855 Wairarapa Fault Symposium: Te Papa Ton- Grapes, R.H., 1999, Geomorphology of faulting: The Waira- Costa Rica: Geological Society of America Bulletin, garewa, Museum of New Zealand, p. 21–30. rapa fault, New Zealand: Zeitschrift für Geomorpholo- v. 107, p. 463–473, doi: 10.1130/0016-7606(1995)107<0463: Berryman, K.R., 1987, Tectonic processes and their impact on gie, v. 115, p. 191–217. QUASCD>2.3.CO;2. the recording of relative sea-level changes, in Devoy, Grapes, R.H., and Downes, G., 1997, The 1855 Wairarapa, McCormack, F.G., Hogg, A.G., Blackwell, P.G., Buck, C.E., R.J.N., ed., Sea Surface Studies—A Global View: Lon- New Zealand, earthquake—Analysis of historical Higham, T.F.G., and Reimer, P.G., 2004, SHCal04 south- don, Croom Helm, p. 127–161. data: Bulletin of the New Zealand National Society of ern Hemisphere calibration 0–1000 cal BP: Radiocar- Berryman, K.R., 1990, Late Quaternary movement on the Wel- Earthquake Engineering, v. 30, p. 271–369. bon, v. 46, p. 1087–1092. lington fault in the Upper Hutt area, New Zealand: New Grapes, R.H., and Downes, G.L., 1999, The 1855 Wairarapa, McSaveney, M.J., Graham, I.J., Begg, J.G., Beeu, A.G., Hull, Zealand Journal of Geology and Geophysics, v. 33, New Zealand, earthquake—historical data: Institute of A.G., Kim, K., and Zondervan, A., 2006, Late Holocene p. 257–270. Geological and Nuclear Sciences Science Report, v. 99, uplift of beach ridges at Turakirae Head, south Welling- Berryman, K.R., 1993, Age, height and deformation of Holo- no. 16, 267 p. ton coast, New Zealand: New Zealand Journal of Geol- cene marine terraces at Mahia Peninsula, Hikurangi Grapes, R.H., and Wellman, H.W., 1988, The Wairarapa Fault: ogy and Geophysics, v. 49, p. 337–358. subduction margin, New Zealand: Tectonics, v. 12, Victoria University of Wellington, Geology Board of Merritts, D.J., 1996, The Mendocino triple junction: Active p. 1347–1364, doi: 10.1029/93TC01542. Studies, v. 4, 54 p. faults, episodic coastal emergence, and rapid uplift: Bookhagen, B., Echtler, H.P., Melnick, D., Strecker, M.R., and Heron, D., Van Dissen, R., and Sawa, M., 1998, Late Quaternary Journal of Geophysical Research, v. 101, no. B3, Spencer, J.Q.G., 2006, Using uplifted Holocene beach movement on the Ohariu fault, Tongue Point to MacKays p. 6051–6070, doi: 10.1029/95JB01816. berms for paleoseismic analysis on the Santa Marıa Crossing, North Island, New Zealand: New Zealand Jour- Moore, P.R., 1987, Age of raised beach ridges at Turakirae Island, south-central Chile: Geophysical Research Let- nal of Geology and Geophysics, v. 41, p. 419–439. Head, Wellington: A reassessment based on new radio- ters, v. 33, p. L15302, doi: 10.1029/2006GL026734. Hicks, S.R., and Woodward, D.J., 1978, Gravity models of the carbon dates: Journal of the Royal Society of New Zea- Burbank, D.W., and Anderson, R.S., 2001, Tectonic Geomorphol- Wairarapa region, New Zealand: New Zealand Journal land, v. 17, p. 313–324. ogy: Malden, Massachusetts, Blackwell Science, 273 p. of Geology and Geophysics, v. 21, p. 539–544. Mouslopoulou, V., Nicol, A., Little, T.A., and Walsh, J.J., 2007, Cape, C.D., Lamb, S.H., Vella, P., Wells, P.E., and Woodward, Hinton, A.C., and McSaveney, M.J., 2007, Lithology and Displacement transfer between intersecting regional W.J., 1990, Geological structure of the Wairarapa Val- hardness of shore platforms and surfi cial boulders at strike-slip and extensional fault systems: Journal of ley, New Zealand, from seismic refl ection profi ling: Turakirae Head, New Zealand: New Zealand Journal of Structural Geology, v. 29, no. 1, p. 100–116, doi: 10.1016/ Journal of the Royal Society of New Zealand, v. 20, Geology and Geophysics, v. 50, p. 27–32. j.jsg.2006.08.002. p. 85–105. Hull, A.G., and McSaveney, M.J., 1996, A 7000 Year Record of Neal, A., Pontee, N., and Pye, K., 2002, Internal structure of Carter, R.W.G., and Orford, J.D., 1993, The morphodynam- Great Earthquakes at Turakirae Head, Wellington, New mixed sand-and-gravel beach deposits revealed using ics of coarse clastic beaches and barriers: A short- and Zealand: Institute of Geological and Nuclear Sciences ground-penetrating radar: Sedimentology, v. 49, p. 789– long-term perspective: Journal of Coastal Research, Client Report 33493B.10, 14 p. 804, doi: 10.1046/j.1365-3091.2002.00468.x. v. 15, p. 158–179. Kelsey, K.M., and Cashman, S.M., 1995, Structural evolu- Neal, A., Richards, J., and Pye, K., 2003, Sedimentology of Cisternas, M., Atwater, B.F., Torrejón, F., Sawai, Y., Machuca, tion along the inner forearc of the obliquely conver- coarse-clastic beach-ridge deposits, Essex, southeast G., Lagos, M., Eipert, A., Youlton, C., Salgado, I., Kama- gent Hikurangi margin, New Zealand: Tectonics, v. 14, England: Sedimentary Geology, v. 162, no. 3–4, p. 167– taki, T., Shishikura, M., Rajendran, C.P., Malik, J.K., Rizal, p. 1–18, doi: 10.1029/94TC01506. 198, doi: 10.1016/S0037-0738(03)00136-2. Y., and Husni, M., 2005, Predecessors of the giant 1960 Kennedy, D., 2008, Recent and future higher sea levels in Nelson, A.R., and Manley, W.F., 1992, Holocene coseismic Chile earthquake: Nature, v. 437, p. 404–407, doi: 10.1038/ New Zealand: A review: New Zealand Geographer, and aseismic uplift of Isla Mocha, south-central Chile: nature03943. v. 64, p. 105–116, doi: 10.1111/j.1745-7939.2008.00132.x. Quaternary International, v. 15–16, p. 61–76, doi: Cochran, U., Hannah, M., Harper, M., Van Dissen, R., Berry- Lamarche, G., 2005, Long-term slip rates and fault inter- 10.1016/1040-6182(92)90036-2. man, K., and Begg, J., 2007, Detection of large, Holocene actions under low contractional strain, Wanganui Nicol, A., Van Dissen, R., Vella, P., Alloway, B., and Melhuish, earthquakes using diatom analysis of coastal sedimen- Basin, New Zealand: Tectonics, v. 24, p. TC4004, doi: A., 2002, Growth of contractional structures during the tary sequences, Wellington, New Zealand: Quaternary 10.1029/2004TC001699. last 10 m.y. at the southern end of the Hikurangi forearc Science Reviews, v. 26, no. 7–8, p. 1129–1147, doi: Langridge, R., Van Dissen, R., Cochran, U., Litchfi eld, N.J., Ber- basin, New Zealand: New Zealand Journal of Geology 10.1016/j.quascirev.2007.01.008. ryman, K.R., Begg, J.G., Villamor, P., Heron, D., Nicol, A., and Geophysics, v. 45, p. 365–385. Darby, D.J., and Beanland, S., 1992, Possible source models and Townsend, D., 2005, Active faulting and paleoearth- Nicol, A., Mazengarb, C., Chanier, F., Rait, G., Uruski, C., and Wal- for the 1855 Wairarapa earthquake, New Zealand: Jour- quakes in the Wairarapa and Wellington regions, in Lan- lace, L., 2007, Tectonic evolution of the active Hikurangi nal of Geophysical Research, v. 97, no. B9, p. 12,375– gridge, R., Townend, J., and Jones, A., eds., Proceedings subduction margin, New Zealand, since the Oligocene: 12,389, doi: 10.1029/92JB00567. Volume: The 1855 Wairarapa Fault Symposium: Te Papa Tectonics, v. 26, p. TC4002, doi: 10.1029/2006TC002090. Darby, D.J., and Beavan, J., 2001, Evidence from GPS mea- Tongarewa, Museum of New Zealand, p. 49–65. Orford, J.D., Carter, R.W.G., Jennings, S.C., and Hinton, A.C., surements for contemporary interplate coupling on Lee, J.M., and Begg, J.G., 2002, Geology of the Wairarapa Area: 1995, Processes and timescales by which a coastal the southern Hikurangi subduction thrust and for par- Lower Hutt, New Zealand, Institute of Geological and gravel-dominated barrier responds geomorphically to titioning of strain into the upper plate: Journal of Geo- Nuclear Sciences Map Series 11, 66 p., scale 1:250,000. sea-level rise: Story Head barrier, Nova Scotia: Earth physical Research, v. 106, no. B12, p. 30,881–30,891, doi: Lensen, G.J., and Vella, P., 1971, The Waiohine River faulted Surface Processes and Landforms, v. 20, p. 21–37, doi: 10.1029/2000JB000023. terrace sequence: Bulletin of the Royal Society of New 10.1002/esp.3290200104. DeMets, C., Gordon, R.G., Argus, D.F., and Stein, S., 1990, Zealand, v. 9, p. 117–119. Ota, Y., and Yamaguchi, M., 2004, Holocene coastal uplift in Current plate motions: Geophysical Journal Interna- Litchfi eld, N., and Berryman, K.R., 2005, Correlation of the western Pacifi c Rim in the context of late Quater- tional, v. 101, p. 425–478, doi: 10.1111/j.1365-246X.1990. fl uvial terraces within the Hikurangi margin, New nary uplift: Quaternary International, v. 120, p. 105–117, tb06579.x. Zealand: Implications for climate and base level con- doi: 10.1016/j.quaint.2004.01.010. DeMets, C., Gordon, R.G., Argus, D.F., and Stein, S., 1994, trols: Geomorphology, v. 68, p. 291–313, doi: 10.1016/ Ota, Y., Williams, D.N., and Berryman, K., 1981, Late Quater- Effect of recent revisions to the geomagnetic rever- j.geomorph.2004.12.001. nary Tectonic Map of New Zealand, Sheets Q27, R27 sal time scale on estimates of current plate motions: Litchfi eld, N., Ellis, S., Berryman, K.R., and Nicol, A., 2007, and R28: Wellington, with Notes: Lower Hutt, New Zea- Geophysical Research Letters, v. 21, p. 2191–2194, doi: Insights into subduction-related uplift along the land Department of Scientifi c and Industrial Research, 10.1029/94GL02118. Hikurangi margin, New Zealand, using numerical scale 1:50,000. Engels, S., and Roberts, M.C., 2005, The architecture of pro- modeling: Journal of Geophysical Research, v. 112, Pillans, B., and Huber, P., 1995, Interpreting coseismic defor- grading sandy-gravel beach ridges formed during the p. F02021, doi: 10.1029/2006JF000535. mation using Holocene coastal deposits, Wellington, last Holocene highstand: Southwestern British Colum- Little, T.A., Schermer, L., Van Dissen, R., Begg, J., and Carne, R., New Zealand: Quaternary International, v. 26, p. 87–95, bia, Canada: Journal of Sedimentary Research, v. 75, 2008, Field Trip 5: Southern Wairarapa Fault and Whareka- doi: 10.1016/1040-6182(94)00049-B. p. 1052–1064, doi: 10.2110/jsr.2005.081. uhau Thrust (Palliser Bay), in Geological and Geophysical Plafker, G., 1969, Tectonics of the March 27, 1964, Alaska, Ferranti, L., Monaco, C., Antonioli, F., Maschio, L., Kershaw, Societies of New Zealand, “GeoSciences ’08” Confer- Earthquake: U.S. Geological Survey Professional Paper S., and Vladimiro, V., 2007, The contribution of regional ence (Wellington, New Zealand): Geological Society of 543-I, 174 p. uplift and coseismic slip to the vertical crustal motion New Zealand Miscellaneous Publication 124B, p. 75–120, Plafker, G., and Rubin, M., 1978, Uplift history and earthquake in the Messina Straits, southern Italy: Evidence from http://www.victoria.ac.nz/geosciences08/fi eldtrips.html. recurrence as deduced from marine terraces on Middle- raised late Holocene shorelines: Journal of Geophysical Lomnitz, C., 2004, Major earthquakes of Chile: A historical ton Island, Alaska: U.S. Geological Survey Open-File Research, v. 112, p. B06401, doi: 10.1029/2006JB004473. survey, 1535–1960: Seismological Research Letters, Report 78-943, p. 687–722. Ghani, M.A., 1978, Late Cenozoic vertical crustal movements v. 75, p. 365–378. Plafker, G., LaJoie, K.R., and Rubin, M., 1992, Determining in the southern North Island, New Zealand: New Zea- Lyell, C., 1868, Principals of Geology or the Modern Changes of recurrence intervals of great subduction zone earth- land Journal of Geology and Geophysics, v. 21, p. 117. the Earth and Its Inhabitants Considered as Illustrative of quakes in southern Alaska by radiocarbon dating, in Gibb, J.G., 1986, A New Zealand regional Holocene eus- Geology, tenth edition: London, John Murray. Taylor, R.E., Long, A., and Kra, R.S., eds., Radiocarbon tatic sea-level curve and its application to determina- Marden, M., and Neal, V.E., 1990, Dated Ohakean terraces After Four Decades: An Interdisciplinary Perspective: tion of vertical tectonic movements: A contribution to offset by the Wellington fault near Woodville, New Zea- New York, Springer-Verlag, p. 436–453.

LITHOSPHERE | Volume 1 | Number 1 | www.gsapubs.org 27

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021 LITTLE ET AL.

Ramsay, B.C., 2005, Improving the resolution of radiocarbon North Island: New Zealand Journal of Geology and Wellman, H.W., 1967, Tilted marine beach ridges at Cape Tura- dating by statistical analysis, in Levy, T.E., and Higham, Geophysics, v. 47, p. 107–122. kirae, N.Z.: Journal of Geosciences, Osaka City Univer- T.F.G., eds., The Bible and Radiocarbon Dating: Archae- Sibson, R.H., 2006, Charles Lyell and the 1855 Wairarapa earth- sity, v. 10, p. 123–129. ology, Text and Science: London, Equinox, p. 57–64. quake in New Zealand: Recognition of fault rupture Wellman, H.W., 1969, Tilted marine beach ridges at Cape Tura- Reyners, M., 1998, Plate coupling and the hazard of large sub- accompanying an earthquake: Seismological Research kirae, New Zealand: Tuatara, v. 17, p. 82–93. duction thrust earthquakes at the Hikurangi subduction Letters, v. 77, no. 3, p. 358–363, doi: 10.1785/gssrl.77.3.358. Wilson, K., Berryman, K., Cochran, U., and Little, T., 2007a, A zone, New Zealand: New Zealand Journal of Geology Stuiver, M., and Polach, H.A., 1977, Discussion: Reporting of Holocene incised valley inﬁ ll sequence developed on a and Geophysics, v. 41, p. 343–354. 14C Data: Radiocarbon, v. 19, no. 3, p. 355–363. tectonically active coast: Pakarae River, New Zealand: Roberts, E., 1855, Memorandum on the earthquake in the Tompkins, J., 1987, Late Quaternary pollen stratigraphy, geol- Sedimentary Geology, v. 197, no. 3–4, p. 333–354, doi: islands of New Zealand, January 23, 1855, in Taylor, ogy, and soils of an area south of Greytown, Wairarapa, 10.1016/j.sedgeo.2006.10.007. R.G., ed., Te Ika a Maui, or New Zealand and its Inhab- New Zealand: Geological Society of New Zealand News- Wilson, K., Berryman, K., Cochran, U., and Little, T., 2007b, itants: London, Werteim and MacIntosh, Appendix F. letter, v. 77, p. 56. Holocene coastal evolution and uplift mechanisms of the Rodgers, D.W., and Little, T.A., 2006, World’s largest co-seis- Van Dissen, R.J., and Berryman, K.R., 1996, Surface rupture northeastern Raukumara Peninsula, North Island, New mic strike-slip offset: The 1855 rupture of the Wairarapa earthquakes over the last ~100 years in the Welling- Zealand: Quaternary Science Reviews, v. 26, no. 7–8, fault, New Zealand, and implications for displacement/ ton region, New Zealand, and implications for ground p. 1106–1128, doi: 10.1016/j.quascirev.2007.01.005. length scaling of continental earthquakes: Journal of shaking hazard: Journal of Geophysical Research, Yamaguchi, M., and Ota, Y., 2004, Tectonic interpretations of Geophysical Research, v. 111, p. B12408, doi: 10.1029/ v. 101, p. 5999–6019, doi: 10.1029/95JB02391. Holocene marine terraces, east coast of Coastal Range, 2005JB004065. Wallace, L., Beavan, J., McCaffrey, R., and Darby, D., 2004, Taiwan: Quaternary International, v. 115–166, p. 71–81. Schermer, L., and Little, T.A., 2006, Late Quaternary defor- Subduction zone coupling and tectonic block rotation in mation along the Wairarapa fault, North Island New the North Island, New Zealand: Journal of Geophysical Zealand: Annual Meeting of the American Geophysical Research, v. 109, no. B12, doi: 10.1029/2004JB003241. Union, San Francisco, CA: Eos (Transactions, American Wang, N., and Grapes, R., 2007, Infrared-stimulated lumi- Geophysical Union), v. 87, fall meeting supplement, abs. nescence dating of late Quaternary aggradation sur- MANUSCRIPT RECEIVED 23 JULY 2008 T21B–0412. faces and their deformation along an active fault, REVISED MANUSCRIPT RECEIVED 14 OCTOBER 2008 Schermer, E.R., Van Dissen, R., Berryman, K.R., Kelsey, H.M., southern North Island of New Zealand: Geomor- MANUSCRIPT ACCEPTED 15 OCTOBER 2008 and Cashman, S.M., 2004, Active faults, paleoseismol- phology, v. 96, no. 1/2, p. 86–104, doi: 10.1016/j.geo- ogy, and historical fault rupture in northern Wairarapa, morph.2007.07.016. Printed in the USA

28 www.gsapubs.org | Volume 1 | Number 1 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/1/1/4/3044036/i1941-8264-1-1-4.pdf by guest on 28 September 2021